More DD on Snapchat+ adoption after mandatory pay-for-storage changes
I received many DMs after my [first post](https://www.reddit.com/r/wallstreetbets/comments/1qnsqug/i_manually_collected_n13196_datapoints_to_see_how/) asking for an update on the Snapchat subscription data I collected, so here it is.
**BACKGROUND:**
Snapchat recently announced that users with more than 5GB of photos saved would have to start paying a monthly subscription fee. In-app prompting shows users how many friends they have who subscribe to Snapchat+. So I compiled data from 80 respondents (37k+ friends in total) to estimate Snap’s earning potential.
**THE GOOD NEWS:**
I remain confident that subscription-based revenue grew by more than 2x over the past quarter, which should result in \~12.5% growth in run-rate revenue and \~25% bump in enterprise value expectation (though stock price will obviously also be affected by expectations of future signups and whatnot)
**THE BAD NEWS:**
The vast majority of additional value due to new Snapchat+ subs will not be realized as Q4 revenue, but will instead build rapidly over the next year. This is because of **(1)** steep first-month signup discounts, **(2)** in-app prompting beginning in waves throughout the quarter, and **(3)** the fact that “Subscription revenue is initially deferred and then recognized ratably over the term of the subscription” (from 10-Q). (That last bit means that revenue from annual subscribers won’t be recognized in Q4 despite Snap receiving the money in Q4).
Also, Apple’s royalty on subscription revenue drops by half after the end of the first year.
**METHOD:**
(See Images 3 and 4)
I broke down user count by age group and location (US, EU, INDIA/ROW) using DAU reports from the 10Q. I matched this with the unweighted average subscription rate for those demographics. I multiplied the two matrices together, summed to get a demographic-weighted average of DAUs who are subscribed, then multiplied by total DAUs. I also multiplied by a “Zombie factor” to account for inactive users on respondents’ friend lists.
This leads me to the 100% increase figure with assumptions:
* Of the average user’s friend list, 40% are DAU, 40% are MAU, and 20% are inactive.
* People over age 35 (making up 30% of the user base) are completely ignored.
* $2/mo subscription plans are completely ignored.
**Assumption analysis**
INDIA:
India accounts for \~40% of users, so their adoption rate carries a lot of weight. I spent 5 hours on Azar (like Omegle but more popular in India), but I couldn’t get a single Indian user to give me their data (though the number one reason for this was that the person already had Snapchat+). So I assumed the Indian Subscription rates as 1/10 of those in the US for the same age group. I think this is reasonable because Snapchat+ is 50 Rupees per month in India, corresponding to 50 cents by exchange rate but $2.50 by Purchasing Power Parity.
DOUBLE-COUNTING:
I used raw averages of % of friends subscribed (as opposed to the friend-count-weighted method I used in my first post). **NOTE:** Correlation coefficient between friend count and % of friends who are subscribed is pretty low (0.29), so this may not be necessary.
INACTIVE USERS:
Inactive users are a major source of error for my analysis. Given that MAUs are 2x higher than DAUs , I think it is reasonable to multiply the measured % by at least 2x to turn it into Snap+ per DAU metric. Some of you are going to say this is unreasonable because inactive users may have fewer friends than DAUs. As a 21-year old, I can anecdotally vouch for a few forces that fight this: A good portion of power users with over 1M snap score make new accounts because having over 1M snap score is sometimes seen as an indicator of promiscuity. Also, criminals love Snapchat for self-deleting texts. Both of these user groups are likely to have inactive secondary accounts with tons of friends.
**OTHER NOTES:**
I removed all respondents who reported 0 Snap+ friends. I attribute this to an error in sampling; these users assumed that, because they were sent to a different page than was shown, their Snap+ friend count must be 0. This was directly reported by a few people who revised their comments on my previous post as well as seen during my manual collection.
Snap doesn’t report raw Subscriber numbers in the 10Q, instead reporting them in the earnings call afterwards.
I won’t post my current position value because my options are so illiquid. Schwab was showing me +$19k in (totally fake) unrealized P&L on Tuesday based on mid-to-market rates (the market maker increased the lowest ask on some of my options by a factor of 70x). If anyone knows how to change this, please comment below.
**Safe to say I’m down a lot. But I haven’t sold anything.**
Holding till Feb 4th. Cross your fingies for me.
PS: See the raw data [Here](https://docs.google.com/spreadsheets/d/1whqLZUYDmS6JkbEILMCOFt6_eUqhgNPnQgHRy2w-Yhg/edit?usp=sharing). I'd love to see your analysis!
PPS: If you know any 35-yr-olds or Indians who use Snapchat, please ask them for their data and comment it below.