Growth Rates: Predicting Growth made easy

This is part 5 of our series on Growth Rates, previous segments are available in our archives.

Almost immediately after you define your growth metric, you will want to project it into the future. Will you grow more quickly this year? How long will your current growth rate continue? The better you can predict your future growth, the more accurately you can allocate your resources and plan ahead.

Unfortunately, real world growth can be hard to predict. For example, given the following daily revenue data let us try and predict the revenue for each day next week:

gdiag-01

Like most real world data there are patterns and trends hidden in this data, making extrapolation difficult. Luckily, there is an easy way to model the growth of cyclical weekly data. Instead of trying to understand the data as a whole, we can observe that the repeating cycle means we can focus on each day of the week independently. That way, way can use a technique like Linear Regression to give us a prediction for every day next week. [1]

For example, if we just analyze the Mondays, the trend is actually a straight line!

gdiag-02

By creating a separate trend for each day of the week, we can build a model for what we think revenue will be every day next week with a fairly high degree of accuracy (we could use more advanced techniques like Double Exponential Smoothing to smooth out the individual estimates):

gdiag-03

With this projection in hand, we can calculate our future growth rates in the same way we have in previous chapters this week!

 

In Review: Whether your business is growing or not is an important fact, but how fast it is growing can be hard to nail down. The way you calculate and predict your growth will depend on how you define growth for your business and is a decision best made early.

 

Next Week: Recommendation systems are all around us. They influence the movies we watch, the products we buy and the trips we take. However, recommendation systems are not just useful for recommending products, they can be useful tools in making decisions. We’ll review how recommendation systems work and how to put them to work for you.

 

[1] If this looks familiar, it’s because we covered it previously on our series on Predicting the Future. Check out the archives to read more.

 

Quote of the Day: “I may be surprised. But I don’t think I will be.” ― Andrew Strauss