Wrote a detailed over of $RDDT earnings and where I think the opportunity is for Reddit. Read here:
seekingalpha.com/article/496…
Which personal agent are you finding the most useful?
31%Grok Bot
42%Muse
27%Other (comment)
26 votes • Final results 596
Today may ultimately be a good event for non-semi exposed stocks:
GDP is doing quite well and overall businesses are growing.
But all of the recent S&P gains have gone into semiconductors. The market had allocated a record 18% or so of S&P 500 into this sector.
This has sucked the oxygen out of the rest of the market. Non-semi stocks are at 2-4 year lows.
If the market believes semiconductors should get a higher risk discount, eventually, that money will rotate elsewhere.
But everything is selling off today?
Yes, because panic and deleveraging etc... but most of it started and centered around uncertainty in the AI semi trade.
Eventually, that capital will be re-deployed. And if it isn't being deployed at same levels back into semi stocks, it may to be deployed into the rest of the market and those that are less semi-exposed.
Also, the semi businesses may still perform well as a business. The market just may have priced in more risk and waiting for future cash flows to come before assigning the present value to them.
S&P 500 finished the day down 4.35%
Very rare event.
And it's always been accompanied by a 4% up day over the past 25 years.
1
6
1,171
Any updates here @shai_wininger ?
$LMND
$ROOT may be moving more mainstream with customers.
Typically, insurers find a niche and grow it. Like Progressive started with non-standard drivers. AKA, high risk drivers. And then they got better at pricing and moved up-market.
Root now says they have meaningful improvement in competitive pricing for "preferred drivers". This is the "low-risk" bread and butter type driver that everyone competes for.
If Root's pricing is good enough to go after chunks of this segment, it opens up a much larger type of TAM customer. It also means their pricing is likely pretty sophisticated if growth is able to come from preferred.
1
5
1,135
$ROOT said on their quarterly call they just launched their best pricing improvement yet.
Might not filter to top or bottom line immediately as this stuff works through lifetime value, retention, etc... as a function of price. It also helps growth, because it should mean lower prices to new customers that deserve lower prices.
Here's the quote:
"We just launched our most predictive pricing algorithm we have ever launched that also had the most improvement in predictive power of a model over its predecessor"
13
1,454
On the $ROOT quarterly report:
Numbers came in pretty much around where you expect based on previous quarters.
Biggest official news is progress with Hyundai Secure+ launching the pilot, and obviously expecting this to later be taken nationwide.
Once Hyundai rolls out nationwide, that will roll over to additional IFP growth, as well as incremental improvements in their pricing.
They don't really need to do a ton to justify a P/E here of 13.5. That number doesn't include a tax credit coming worth half their current market cap.
5
17
1,888
For $LMND, overall, decent quarter. They made good net income progress when adding back sales spend.
The loss ratio is trending a bit unfavorably, but given that net income adding back S&M continues to improve, it's not really a big deal.
If anything, it would have been a great quarter if not for Pet & Europe GLR.
Management should give a bit more clarity on Europe and breakdown by product. There was a huge swing in GLR, with it going above 100%. They said on the call that it was mostly driven by large losses. Is this primarily a Renters or Home in total IFP product in Europe? US Renters generally has less volatility.
One other note:
Not a fan of the continued focus on employee count instead of opex or automation metrics. Lemonade appears to have non-employees doing the customer service work on claims, as well as other things. Those workers are contractors, and get paid, and work with Lemonade, but they aren't legally considered employees.
This metric is easily abused, and Lemonade could have hired 50% more contractors to handle 50% growth in pet claims for instance.
A better metric for Lemonade to report is either:
1. Opex cost ex-growth spend, could even adjust it further to take out SBC or
2. Some kind of automation or efficiency ratio
Lemonade used to provide occasional tracking from Shai on X of % of customer support tickets handled by AI through resolution. We haven't got an update there in maybe 2 years.
Lemonade also did a lot of work on having AI customer support systems. Would be good to know if that's actually implemented and driving more automation, and if so how much.
If a big chunk of automation is from AI systems, they should report it so we know. Report % of tickets fully handled through AI, through AI phone chats, etc..
Here's the progress of $ROOT & $LMND net income when you add back sales and marketing + commissions, updated for Q3.
$LMND has made some consistent incremental gains, including this quarter.
$ROOT is doing well, but is still holding back direct marketing spend and deploying capital in other ways (buybacks, or saving to deploy later).
Some framing around $RDDT users and their strategy
Reddit US WAU is 188.5 million. That is half the US population.
US DAU is 50 million.
What Reddit is trying to do is convert those Weekly users into DAU.
The needed users are already returning weekly. Creating a daily habit from those users is where the DAU growth will come
As Huffman has been saying, they want to turn them into Redditors.
That is why they are going to force logging into the app. This is about getting those WAU into logged in WAU. And later logged in DAU.
Once they are logged-in, they can then give those WAU users custom feed recommendations and convert more of them into DAU. Get more of them to use Reddit search. Customize and tailor the experience for retention, etc...
It's not about saying, "referral traffic isn't helpful" so we will force logins.
It's that, we want WAU to become DAU, and requiring logins will help with that. Among other things.
In that process, Reddit could lose some MAU or WAU, but those are low value users. DAU is the high value users who spend on average 30 minutes a day on the platform.
Forced logins actually increase DAU, in that way.
All other social media networks already force logins, so they can deliver custom feeds as well.
$RDDT Summary and Thoughts on Q3'2026
The user news
- US DAU down 4.7% sequentially
- DAU could fall around 10-12% (low double digit) in Q4 from Google traffic
- Around 80% of monetization comes from direct/logged-in traffic.
If logged-out traffic drives roughly 20% of revenue, a 10% decline in that traffic would mean roughly 2% less revenue.
Numbers for Q3 are:
- Revenue up 53% YoY
- Net income up 74% YoY
- P/E of 21.8
- Q4 user headwind would be about 2% less revenue
A 2% revenue headwind against 53% growth is relatively small.
The question the market has is: Does Reddit keep its direct user base with less referral traffic?
For Reddit, it's an unknown. We don't get that answer except one quarterly report at a time.
It's up to each investor to decide how much of a risk penalty to assign to less referral traffic. From a revenue/net income perspective, there is minimal penalty this year, next, or the year after. They also have all that time to find new ways to generate more revenue/users/etc... or for DAU to find a base with referral and direct.












