Unfortunately, it seems we have made a mistake with one of the two models which we used to calculate the number of domains churning off. We have discovered this mistake in the early hours of June 29. 2018 We would hereby like to make a correction of the second churn model and show how it affects the calculations in the report. We apologize for the inconvenience.
We could have stayed silent about this error (especially because it doesn’t change much in our thesis) but our upbringing and conscience didn’t permit this.
These are the key things that change:
- In the original report we claimed that, using the second method to calculate churn, 93.3% of domains 1 year or younger churn off. We now believe the right number is 78.97%.
- If we combine with the first churn method (looking at churn through expiring domains), which doesn’t change, we now think that the percentage of clients that die before they reach their first anniversary is 77%, not 85% as we stated in the original report. 77% is still a very significant number.
- Because of this change, we now believe that the average customer lifetime is not 14 months, but 15.58 months. Still a very low number in comparison to other SaaS companies.
- Since we were overcounting churned domains, we now think that Shopify is still adding more domains than losing in the .com realm. In the previous report we alluded to the fact that domains added are starting to be very close or almost equal to domains lost
- The lengthening of the customer lifetime slightly affects the Dollar-based retention rates we calculated for Shopify.
- There are some changes in the analysis of the Weebly zone file as well. We claimed that Shopify’s number of domains churning to the total number of domains hosted is about 4.3x more than at Weebly for a similar number of domains hosted. We now believe this number to be 2.57x.
- Because of the slight increase in Shopify customer lifetimes we raise our price target from $60 to $65
It is important to understand that these changes only affect the first part of our report that deals with Shopify customer churn and Dollar-based retention rates. The other sections that deal with Shopify’s operating leverage, top250 clients, health of Shopify’s customer base, Facebook, Shopify Capital and Shopify Plus are unchanged.
We would also like to say something about how this error was made. It is very technical. Essentially, we discovered that a very small subset of domains hosted by Shopify (and Weebly) have wrongly set DNS records. Essentially, they have their original DNS settings and new DNS settings related to Shopify as well. This causes the domain to resolve to Shopify’s IP address on some days and to another IP address on another day, for example. So, this artificially increased the number of domains churning off. With this revision of this report we have now gotten rid of this misleading effect.
Now we would like to show how some of the tables we presented change after this revision. We always include the former table and the new table after the corrections have been made so you do not have to look back and forth between the reports.