1 00:00:00,510 --> 00:00:02,500 As we discussed in the last video, 2 00:00:02,500 --> 00:00:07,340 depending on who you talk to churn calculations should either include or 3 00:00:07,340 --> 00:00:11,260 exclude new customers you acquire over the timeframe you're analyzing. 4 00:00:12,290 --> 00:00:16,970 Some people refer to this distinction as gross churn versus net churn. 5 00:00:16,970 --> 00:00:19,520 I like that way of thinking about it because it's clear 6 00:00:19,520 --> 00:00:21,200 what the distinction is. 7 00:00:21,200 --> 00:00:25,340 Gross churn doesn't include new customers, net churn does. 8 00:00:25,340 --> 00:00:27,810 Let's walk through a few examples together. 9 00:00:27,810 --> 00:00:33,500 If we started the year with 100 customers and 20 customers churn out, and 10 00:00:33,500 --> 00:00:38,745 we added 30 new customers over the year, our gross customer churn rate would be 11 00:00:38,745 --> 00:00:43,700 20%, the 20 divided by the 100 beginning of year customers number. 12 00:00:43,700 --> 00:00:48,320 But if we take into consideration adding 30 new customers, 13 00:00:48,320 --> 00:00:50,510 our net churn rate will be lower. 14 00:00:50,510 --> 00:00:53,130 Let's look at how different our churn number will be 15 00:00:53,130 --> 00:00:57,640 if we added our acquired customers to the numerator or denominator. 16 00:00:57,640 --> 00:00:58,920 It's a significant difference. 17 00:01:00,250 --> 00:01:04,200 One is still positive, and the other is actually a negative percentage. 18 00:01:04,200 --> 00:01:07,370 So here we have the new customers 19 00:01:07,370 --> 00:01:10,610 adding to the denominator of our churn calculation and 20 00:01:10,610 --> 00:01:16,400 below we have the new customers added into the numerator of our churn calculation. 21 00:01:16,400 --> 00:01:20,800 When calculating the number in the numerator, this metric is more useful for 22 00:01:20,800 --> 00:01:25,405 signaling whether or not our customer base grew or shrank over the timeframe. 23 00:01:26,530 --> 00:01:30,360 If it's a negative church, our customer base expanded, if positive churn, 24 00:01:30,360 --> 00:01:31,730 it declined. 25 00:01:31,730 --> 00:01:34,670 When we are looking at our customer base overall at Treehouse, 26 00:01:34,670 --> 00:01:37,380 we include new students in the denominator. 27 00:01:37,380 --> 00:01:41,620 However, and apologize for complicating things a little bit, we look at 28 00:01:41,620 --> 00:01:45,560 different calculations for churn depending on the question we are trying to answer or 29 00:01:45,560 --> 00:01:47,930 decision we are trying to make. 30 00:01:47,930 --> 00:01:50,910 I hope by now you are starting to appreciate 31 00:01:50,910 --> 00:01:55,400 how the way we calculate churn can have some large implications. 32 00:01:55,400 --> 00:01:59,280 Indeed this metric is not without controversy. 33 00:01:59,280 --> 00:02:03,145 Netflix used to report subscriber churn in their financial statements and 34 00:02:03,145 --> 00:02:06,590 stopped several years ago as they asserted that 35 00:02:06,590 --> 00:02:09,850 the churn metric is a less reliable measure of business performance. 36 00:02:10,990 --> 00:02:14,450 Years before that, they were reportedly sued by some shareholders for 37 00:02:14,450 --> 00:02:15,790 calculating churn incorrectly. 38 00:02:16,800 --> 00:02:19,880 Again, as there is no standardized formula, 39 00:02:19,880 --> 00:02:24,220 make sure you are calculating something that is useful for managing the business. 40 00:02:25,300 --> 00:02:29,880 I personally prefer to exclude new customers from churn calculations because 41 00:02:29,880 --> 00:02:33,490 I am of the view that churn is most valuable as a reflection 42 00:02:33,490 --> 00:02:38,670 of our existing customers' behavior, but this is far from a universal practice. 43 00:02:38,670 --> 00:02:43,850 Again, as is often the case in the world of data analysis, we need to exercise 44 00:02:43,850 --> 00:02:48,650 our analytical skills to decide what makes the most sense for the situation at hand. 45 00:02:49,790 --> 00:02:53,450 It's good to explore if there are norms for the industry you're in. 46 00:02:53,450 --> 00:02:56,910 Also note that as your business changes so 47 00:02:56,910 --> 00:03:00,520 might the way you think about churn and or retention. 48 00:03:00,520 --> 00:03:04,440 Perhaps a calculation that is appropriate when you are in exponential growth mode 49 00:03:04,440 --> 00:03:08,270 doesn't make sense anymore when your revenue growth slows down. 50 00:03:08,270 --> 00:03:09,130 There is nuance. 51 00:03:10,430 --> 00:03:14,120 My favorite way to look at churn is based on revenue. 52 00:03:14,120 --> 00:03:18,790 In many situations, not all customers pay the same amount each month. 53 00:03:18,790 --> 00:03:22,310 So we'll bring back the concept of MRR and 54 00:03:22,310 --> 00:03:24,770 walk through a dollar-based example in the next video.