Andrew Freedman, CFA 🦅

To each their own… But IMO you can’t judge a business that is growing/scaling as rapidly as $ANT based on sequential trends… especially when you ramp compute ahead of revenue/product launches. The trailing or YoY trend in margins is accurate reflection of scale. Also there is going to be lumpiness around model training cycles and product launches. Now if incremental margins continue to deteriorate over a period of multiple quarters or start declining YoY that would be a problem. Or if operating losses grow as revenue base gets larger that is a problem. We also knew that the short term compute deal from $SPCX was coming online… and that compute expense was going to ramp. So again, timing mismatch. And looks like a lot of that going to training (but I need to look closer). Point is, I don’t understand how you can read anything structurally negative about this business from Q3 vs Q2 margin trajectory alone… To each their own, but I think looking at the blended Q2/Q3 incremental margin is probably a better way to think about scale and where margins are trending. And on that basis you get something closer to mid-20s which is still above the Q3 operating margin. Also why I focus so much on the unit economics of the business… because that tells you if there is a structural issue to scale.

From our 4Q26 Themes: $RDDT is losing Google surface area.Reddit's keywords in Google's SERP features fell 17.5% from June and 31.7% from December. Same-store spend from existing advertisers grew only 7% in 3Q26, even as one agency's count of its advertisers on Reddit rose 65%. We would read a 3Q growth rate in the mid-50s as confirmation. Consensus is 49.6%, and we think buy-side expectations sit closer to 60%.