Correction: I would like to issue a correction from yesterday’s piece on Synopsys. I mentioned the company is trading at ~56x LTM EBITA. The problem is Synopsys’ LTM EBITA doesn’t incorporate Ansys’ EBITA which they just acquired. As a result, while the EV includes the impact of acquisition, my LTM EBITA did not. When I added Ansys’ last four quarters reported EBITA (again, deducted SBC to keep it apple-to-apple), Synopsys is actually trading at ~36x EBITA (assuming $90 Billion EV). While I still am not interested in Synopsys at this valuation, it is not as egregious as I implied yesterday. Thanks to everyone who pointed out the error.

Meta CFO at GS Communicopia

There’s been a lot of discussion about Zuck claiming during the White House meeting that Meta is going to “invest” $600 Billion in the US by 2028, but CFO Susan Li explained that this number includes all the capex and opex in the US. I estimate Meta is probably going to spend $450-500 Billion in opex (ex-depreciation) between 2025 and 2028. If you assume ~75% of this opex will be spent in the US, that leads us to ~$350 Billion. When you add another ~$250 Billion of capex i.e. ~$60 Billion/year in the US between 2025-28 period, you can get to $600 Billion “investments”. That actually seems not as outlandish as it may have sounded when Zuck first said it.

Of course, these are staggering numbers, but Li reminded that they remain committed to operating profit growth:

I believe Mark has publicly committed to delivering operating profit growth. And I realize that, that is not stand-alone a benchmark that is extremely exciting and that in practice, we, in fact, have to make sure that over the long run, we are an attractive investment relative to any of the many other public equity investments that are available to all of you.

So there will be obviously like lumps in the years. It would be a truly amazing thing if you could sort of just deliver nice linear compounding returns in a predictable way forever. But we're committed to making sure that we deliver attractive financial returns over time and across and we sort of think about managing the portfolio of investments in that way.

She also made some interesting comments on the evolution of ad loads. Ben Thompson, for example, has been a bit skeptic and wondered whether Meta intentionally dialed up the ad load in recent times to receive “permission” from investors to go after more speculative investments. That’s a fair concern. From Ben Thompson:

I don’t think it’s a coincidence that, in the same quarter where Meta decided to very publicly up its investment in the speculative “Superintelligence”, users got pushed more Reels and Facebook users in particular got shown more ads. The positive spin on this is that Meta has dials to turn; by the same token, investors who have flipped from intrinsically doubting Meta to intrinsically trusting them should realize that it was the pre-2022 Meta, the one that regularly voiced the importance of not pushing too many ads in order to preserve the user experience, that actually deserved the benefit of the doubt for growth that was purely organic. This last quarter is, to my mind, a bit more pre-determined.

However, Susan Li alluded that their ad load has gotten much more discerning in terms of when to show ads:

"with ad load, it is also a story of personalization and of increasingly trying to infer when you are using our product, when you are in a session, are you interested in buying something? Are you in a commercial state of mind?

there will be times when I'm like clearly scrolling through friends and family content and probably not thinking about shopping, and that's a good time to show me fewer ads. And that enables us without any meaningful sort of engagement impact to really optimize the impressions that we show you and increase the value of those impressions.

And so there's a lot of work that's being done, I think, to really make ad load, it's gone away from kind of like 12.5%, the 1 in every 8 stories as an ad to something that feels really, really sort of tailored to when you are most likely to want to have a commercial experience.

“On the demand side, reported CPM is really an output of the work that we are actually doing to drive prices down… even as we bring that cost down, you should see reported CPMs go up because we are making each impression convert more frequently and be more valuable.”

One inference I had from this is reels may be even more conducive to show you more ads compared to newsfeed. Meta may not be dialing up ad load mindlessly, rather carefully capitalizing on every opportunity of increasing ad loads whenever it doesn’t hurt user engagement. It’s not just ad load personalization, even the ad content itself is likely going to be personalized in not-so-distant future. When I think of Meta’s ads, the path to personalization feels more of asymptote to me…there’s always going to be more and more work to do to personalize our experience even more over time. From Li:

"...the idea that ads can be super tailored to each person without the advertiser knowing that you and I could get the like a hotel in Hawaii could target you and I with the same ad... But we know that you should get an ad that's oriented around like Big Wave surfing and I should get an ad that's oriented around like hiking. And it just creates those for us because it knows that those are our interest and the hotel doesn't have to, we do all that on behalf of the advertiser."

Li also pointed out the paradox of productivity gain from AI; while the concern seems to be around whether companies will hire fewer people because of higher productivity, Li mentioned the opposite should be the case although there is certainly a point at which the marginal return would taper off:

"You can imagine that there are teams like if each of your engineers can produce twice as much product impact because the Al tools have made them twice as productive as they previously were, then we should probably hire a lot more of those engineers. There are sort of other areas around the company where I think this will be more of an efficiency gain..."

"The thing that's a little bit unknown is but like where are you on the sort of marginal curve of returns right now? So sure, maybe we know sort of each 25 engineer unit of work is going to generate some amount of return, but what happens is you add like 1,000 engineers, like how quickly does the curve drop off?"

There was, of course, a lot of discussion on Meta’s investments in Superintelligence team, but there was nothing incremental there. She did mention that Meta AI is being widely used despite the fact that the model powering Meta AI is not a SOTA model. One of my hypotheses is the kind of interaction people likely do with Meta AI doesn’t really need the most advanced SOTA model and Meta may be discouraged to invest ever increasing amount on SOTA models once they realize the incremental engagement gain is not enough to justify to stay in the race. As a result, it may make a lot more sense to trail the SOTA model by 6-18 months and play catch up with a fraction of the cost for SOTA model developers. Of course, even that may not be easy to do as we are finding out Meta’s recent struggles with Llama models. If the Superintelligence team can get Meta back on track, I think that’s what they may focus on and spend much more time in productizing and infusing AI in every facets of the user experience through Meta’s apps.

Microsoft at GS Communicopia

Jared Spataro, Chief Marketing Officer of “AI at Work” at Microsoft, attended GS Communicopia as well. I will share my notes and thoughts on this session as well as OpenAI’s recent partnership with Microsoft behind the paywall.

In addition to "Daily Dose" (yes, DAILY) like this, MBI Deep Dives publishes one Deep Dive on a publicly listed company every month. You can find all the 62 Deep Dives here.