Thread by @borrowed_ideas

1/ Final thread on $ETSY

Lots of thoughtful and gracious feedback (both in support and against) on my decision to sell Etsy. I’ll recap some of these discussions and share my thoughts in this thread.

For the uninitiated, here’s the original post: mbi-deepdives.com/etsy/

2/ Pushback #1: You should look at GMS/active seller instead of GMS/active buyer

If you followed my original work, that’s exactly how I modeled my GMS in future years. Now I believe that’s an inferior approach.

3/ Most marketplace literature comes from VCs who rightly focus on both sellers and buyers in a two-sided marketplace as it faces an acute chicken and egg problem. Since VCs take the marketplace from embryo to a toddler stage, BOTH buyers and sellers are of paramount importance.

4/ But when a marketplace becomes public and/or has been around for more than a decade, public market investors deal with a teen/adult, not toddler.

Sellers start exhibiting power law or pareto distribution. The marginal seller has lot less value to add to the marketplace.

5/ The average seller’s experience on the marketplace can be vastly different from what’s been happening in the right tail of the sellers. Therefore, increase of GMS/active seller has lot less information to convey to the investors.

6/ Once sellers start exhibiting power law distribution, the value of marginal buyer becomes much more important to the marketplace and hence should be, imo, focal point for projections.

~8% habitual buyers at Etsy drives >40% GMS. This is bit of a rorschach test for investors.

7/ Bulls look at that and see potential for converting more active buyers to habitual buyers to drive GMS. Bears see this as sign of general weak health of the marketplace. Based on my new understanding of consistent >50% buyer churn every year, I’m more sympathetic to the bears.

8/ I don’t know the answer, but this could be good area to study/explore further.

What’s the benchmark for top 10%/20%/30% buyer to overall GMS in other two-sided marketplaces that are teen/adult, not toddlers?

9/ Pushback #2: # of Active buyers estimates in out years are too conservative

I sort of expected this pushback because I knew I would do the same a week ago. What bulls are missing is I’m not explicitly forecasting active buyers at all.

10/ Look at the category of buyers closely and it is perhaps more likely all my assumptions by category are aggressive.

In 1Q’21, Etsy mentioned it had 160 mn all-time buyers on the marketplace. I’m modeling another 123 mn new buyers in 2021-2030.

11/ Can you assume 150 mn? Sure, but I would be hard pressed to consider that a base case.

12/ For both re-activated buyers and consistent buyers, my assumptions give Etsy credit as a much better marketplace than it was before pre-Covid since both these assumptions are more lenient to Etsy vs any pre-Covid periods.

13/ Pushback #3: Implied CAC is not dynamic with estimates of active buyers

Here too I believe bulls are underestimating the true nature of CAC for Etsy. What the churn data shows is Etsy essentially owns a lot of “transactions” and very few “buyers/customers”.

14/ Etsy’s CAC applies for all three categories of buyers.

You need to create awareness to acquire new buyer. You need to remind lapsed buyers of Etsy to re-activate them. And you also need to ensure consistent buyers don’t churn off.

15/ I originally thought I bought a two-sided marketplace with network effects that *own* a sizable buyers/customers (after initial churn), not just transactions. I believe there will be tens of other marketplaces/retailers who will and can own transactions.

16/ Terminal value is a big question mark for these marketplaces/retailers. I want to own business that tend to own customers/buyers, and not necessarily on a permanent CAC treadmill.

17/ My guess is Bezos understood AMZN too mostly had “transactions” which is why he launched “Prime” to create “buyers”.

18/ What would change my mind on Etsy?

Hey, I just changed my mind. Maybe it would be a good question for the longs 😉

Just kidding.

19/ In my mind, everything comes down to buyer frequency. It’s unlikely I’m underestimating LT active buyers # by a lot. Etsy needs to own buyers, not transactions. And we’ll know they have buyers if they can increase GMS/active buyers at double digit rate without cranking up CAC

End/ I rate Josh Silverman highly (and whoever wrote the buyer/seller retention section in the 10-K poorly), so possible he can keep making the marketplace better and healthier.

Just don’t feel it’s a good odds anymore looking at the embedded expectations.