A black swan for ecomm

Apple's App Tracking Transparency (ATT) had a huge impact on Facebook (and all social media) ad effectiveness while the ecommerce industry was hit with higher costs and supply constraints.

These all happening at the same time are a great lesson on blow-up risk.

What's a black swan?

Economists have been using the term "black swan" to represent unprecedented events ever since author Nassim Taleb popularized the term.

The black swan theory goes like this:

All swans you have seen in your life have been white. That must mean all swans are white.

But the outlier, one black swan, breaks your entire mental model. The book encourages its readers to never underestimate what we do not know, and to never underestimate the impact of outliers.

So how is ATT, a recession, and supply chain constraints a "black swan"?

ATT is an unknown, entirely from left field. Typical recessions are coupled with lower demand, not this dual sided supply and demand recession.

No one predicted one of these happening. All three appearing at the same time is black swan territory.

Why did we miss it?

Interestingly, the downstream impact of ATT was not anticipated for one-click checkout, mainly because the trend to DTC seemed undeniable. Charts looked up and to the right.

I have been buying from more and more DTC stores, so I 100% believed in this trend.

While ATT did improve privacy for Apple consumers, it was a somewhat non-financial decision and caught the ecommerce industry off guard.

Facebook ads in particular were very vulnerable to this change. Stratechery has a theory that this means Facebook was properly using every data signal they could to make their ads as effective as possible — ATT seems to have proven this.

Facebook's ad efficiency is great for the average DTC company, as they can't employ a large number of data scientists to make their ads as effective as Facebook was doing for them.

Most of these one-click checkout companies have some sort of base fee plus a take rate (percentage of the sale). This revenue model seems highly diversified, because after all, isn't having tons of unique customers considered "diverse"?

At the surface, the answer seems to be "yes", but most DTC suppliers sit somewhere in China, which struggled to keep factories open during Covid. Recently, Zero Covid policies have shut down quite a few factories while the vaccines in the country are proving less effective than their US counterparts.

The entire industry was also very reliant on the efficient Facebook ad machine working, which ATT threw a wrench in.

Things being correlated where they seem diversified is where blow-up risk comes from.

Models likely showed the DTC trend continuing, meaning that their take rate was going to justify large CAC over the long term while serving multiple stores meant diversification.

At the same time, shipping rates went through the roof to get a container of goods to US shores (2k → 20k), increasing the cost of goods for almost every DTC store.

Lower margins meant less money for DTC stores to spend on ads and R&D, which decreased top-of-funnel for 3rd party checkout sites at the exact same time that take rate revenue was being hit due to ATT inefficiencies.

All of these events happening at the same time truly is a black swan event and shows how difficult modeling revenue can be when it's reliant on "the real world".

I think I'll stick to building SaaS and websites for now.