David Perlmutter Profile picture
Apr 24 10 tweets 7 min read Read on X
1/
I'm doing this write-up as a partial response to @realroseceline's post about $ZETA yesterday, linked below. Rose’s post was absolutely top-class, and I admire the writing and thought process very much. If by some strange miracle you are following me and not yet following Rose, stop what you are doing and go follow that account immediately.

Some of my own views on the topic contrast some of Rose’s, so I wanted to write about it. No pressure of course, but if @realroseceline or @wealthmatica, or anyone else for that matter, would like to share thoughts on this, I would welcome it.
2/
First, there's an argument that $ZETA operates (almost) solely at the bottom of the funnel where intent is already at least partially defined. Also up for discussion, the view that Zeta only targets consumers who have previously engaged (at least in some small way) with the client-company.

🌹 “That is not new customer acquisition, it’s called lifecycle marketing. Lifecycle marketing is what happens after someone already exists in your business in some form, whether they are a customer, a visitor, or just someone who engaged once with your email or app.”🌹

🌹“If you take a group of users who already showed intent and improve how you reach them, how often you reach them, and what you show them, the returns will naturally look higher. You are improving conversion on people who were already close to converting, which is a very different task than generating brand new demand from brand new customers.”🌹

$ZETA absolutely does do this, but they should not be put in a box as doing only this. I'm not even sure we can say with reasonable certainty that they do MOSTLY this, though I would forgive speculation that this is the case (since they don't scream it from the rooftops, though I don't personally subscribe to that argument). They do “generate brand new demand from brand new customers.” Stick with me here.
3/
A business that is using ZETA's platform does not have a window of potential customer profiles that is limited in scope to only customers that they have interacted with in some way before. The business’s window of potential customers spans the entirety of the identity graph. Meaning all of the 245M US consumer profiles. That's almost the entire US population. If at any point the stars align data-wise or trigger-wise, any one of these profiles can be the subject of an ad reaching them from the business.

🌹“Now compare that to true new customer acquisition, which is fundamentally a different problem. Instead of organizing known users, you are trying to find new people who have never interacted with your brand, have no data history, and may not even know your product exists. That is not a matching problem, it is a discovery problem that requires scale, data, and constant experimentation.”🌹

If Joe Shmo clicked on a video game ad in an email 3 years ago and bought one, today, when Micro Center becomes a Zeta platform user, Joe may at some point be targeted by Zeta’s platform with an ad for a gaming monitor from Micro Center, even though he has never even heard of the place before. And he may buy it. So, from the perspective of Micro Center, Zeta has acquired a brand new customer for them. That is not traditional top-of-the-funnel acquiring, but from the perspective of the business-client, it looks and feels the same, only more efficient, and with more purposeful targeting.
4/
$ZETA supplies explanations and tutorials for its platform users in The Zeta Knowledge Base. One page in particular gives step by step guides for acquiring brand new customers. The page is titled "Prospect Audiences":

"Prospects – Select this option to segment new audiences from the Zeta Data Cloud, and target them through paid media channels like Display, Audio, and Connected TV.”

Matched lists are also important:

“A matched list is a set of first-party individuals uploaded into the Zeta Marketing Platform, who are then matched to Zeta's Data Cloud and further enriched with potential intender audience membership and other data points. This matched list can then be activated against Zeta's omnichannel solutions or can be extended using our AI-driven lookalike modeling capabilities.”
Source: knowledgebase.zetaglobal.com/kb/prospect-au…Image
5/
The same page describes the lookalike workflow, giving clients a way to answer the following important question:

"Who are the top prospects known to Zeta that are most similar to my seed audience so that I can achieve my goal of reaching X million prospects across channels?” Image
6/
Finally, Zeta describes its audience sources here: knowledgebase.zetaglobal.com/kb/audience-so…

"Zeta Data (Prospects) — Zeta's proprietary second- and third-party data set, which gives you access to millions of anonymized consumer profiles. This source helps you expand beyond your current customer base, discover new prospects, and reach audiences that match your target criteria—across email, display, CTV, social, and more.”

Of course, the data around those profiles, all 245M of them (in the US), and the signals that fly, matter for whether there is a match. And I think this is why people think it's a bottom-of-the-funnel only type of business model. Many of those signals do indeed come from brush-ins with the company, like clicks, or emails, or some other kind of engagement; i.e. the bottom of the funnel. Other signals, however, don’t involve the company at all, yet an ad from the company will still end up going the way of the prospective customer. This acts like a top-of-the-funnel driver for the client.

Zeta does deal in prospecting and new acquisitions, not just lifecycle and remarketing.
7/
Next topic of discussion:

🌹 “There is also a natural ceiling to this model that people don’t think about. You can only optimize the users that exist in your system. If new demand is not constantly being created at the top of the funnel, the opportunity at the bottom eventually slows down.” 🌹

The way $ZETA gets paid is essentially by granting companies access to their first party data/profiles (yes, simplified, but still true). The reason the economics work is because there is no ceiling on the number of clients Zeta can sell this access to. Again, bear with me.

The number of profiles in the graph does not need to grow for Zeta to continue to grow because they will keep acquiring client businesses. The reason I make this point at all is because I think there is a view that, in ZETA's case, top-of-the-funnel acquiring means the number of available consumer profiles expands, i.e. more consumers in the identity graph. But the number of profiles is already almost the whole country. It's a big potential audience in the data graph/profiles, and top-of-the-funnel in this model means growing audience subsets for clients, as I showed above that Zeta actually does through prospecting and putting together look-alike audiences.
8/
Zeta is a connector of two ends. One end is the profile and the other is the business. A ceiling exists if both ends are static in number because you can only optimize the users so much with the businesses you service, and then there is no more upside. If the number of customer profiles goes up, there are more profiles to optimize. If the number of profiles is static at 245M, then you have to increase the number on the other end– the client-businesses. If the number of businesses increases, each customer profile in the graph gains in potential monetizability because each profile now has new potential transactional connections.

If either one of the ends that Zeta works to link up increases, there is no ceiling.
9/
What's interesting is that the more businesses that are using Zeta and have access to the pool of profiles, the more valuable each singular profile becomes. This is because there are more potential transactional touchpoints now per profile. I am not suggesting that every profile will mingle financially with every business that uses $ZETA. But it is true that every new business coming on is potentially a new recipient of a transaction from any given profile, and so each profile is now exposed to more businesses.

More potential business per profile.

That means that with every new business using $ZETA, every profile and the dynamic data attached to it gets incrementally more valuable. There are real economics hidden in that perspective.
10/ End
As a bonus aside, these profiles and the dynamically moving data that are gaining value with each new business coming on board are placed into goodwill when they come from an acquisition. Something to think about as most think of goodwill as a value-sucking line item. But that's another discussion entirely.

I would love any and all thoughts.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with David Perlmutter

David Perlmutter Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @Muzzlebuster

Jan 3
1/5
The Brookfield ecosystem is anything but simple. At its core, it consists of a series of businesses which are connected in different ways. They feed off of each other and are each bolstered by their relationships within the system. Let’s break down the way money flows within the system, particularly between Brookfield Asset Management Ltd $BAM, Brookfield Wealth Solutions Ltd $BNT, and Brookfield Corporation $BN. 🧵👇Image
2/5
Brookfield Asset Management Ltd $BAM earns its money by managing capital from institutional investors, private wealth, and insurance float. BAM deploys the capital into long-term private funds of high quality assets in infrastructure, renewable energy, real estate, private equity, and credit.

Where their money comes from:
BAM raises capital organically from institutions who want to invest with them. The amount of money that BAM manages is called fee-bearing capital because BAM charges management fees as a percentage of this money. The more fee-bearing capital is committed, the more fee-related earnings (FRE) BAM makes.

They also take in fee-bearing capital from Brookfield Wealth Solutions (BWS), the insurance business. The insurance float, which is just premiums paid to them from insurance customers, is trusted to BAM for investing, adding to BAM’s committed pool of capital and, hence, its FRE.

Flywheel element:
As BWS grows its insurance/annuities business, BAM grows in turn.

Finally, BAM makes “realized carried interest” when their investment funds clear hurdles for their investors. This is basically a performance fee for reaching a return goal.

Where their money goes:
Almost all of the FRE translates directly to Distributable Earnings (DE). This is like free cash flow in the Brookfield universe. BAM takes the DE and pays almost all of it out as a dividend to shareholders. Brookfield Corporation $BN is its biggest shareholder (72% of total shares), so they get most of the dividends.

Flywheel element:
As BAM grows FRE, $BN’s earning power grows in turn.

BAM also pays a 33% royalty on all realized carried interest (performance fee from surpassing return goals) to $BN. It retains the rest for redeployment.

Flywheel element:
As BAM’s AUM and fund sizes grow, realized carried interest grows, which in turn grows $BN’s royalties earnings/DE.Image
Image
3/5
Brookfield Wealth Solutions Ltd $BNT (BWS) is the insurance arm of the ecosystem. They operate a reinsurance, insurance, and annuities platform.

Where their money comes from:
BWS takes in premiums or capital from selling retail annuities, taking on institutional annuities, and taking in premiums from Property and Casualty insurance.

Where their money goes:
The capital taken in by BWS is the float, and the float is pushed to BAM for investing. BWS pays BAM a percentage of that capital, which translates to FRE for BAM. BAM makes a return for BWS’s invested float. This return, minus what BWS pays out for insurance claims and other expenses, becomes their profit spread, or Distributable Operating Earnings (DOE).

Though a small dividend is paid out from the DOE, most of it is retained so that BWS can acquire other insurance companies or push more capital towards BAM for management. $BN retains control of these earnings.

Flywheel element:
Brookfield Wealth Solutions grows from selling more annuity/insurance products, acquiring other insurance companies, and reinvesting its DOE which was earned from the spread. As BWS grows, its float grows, which means that BAM’s AUM grows, which means that BAM’s FRE grows, which means that $BN’s distributable earnings grows (dividends from BAM and BWS)Image
Image
Read 6 tweets
Dec 22, 2025
1/10
$GE Aerospace was officially spun-off in April 2024 as the last part of a massive corporate restructuring. The ticker GE gives the investor a pureplay position in the aerospace engines business.

Chris Hohn of TCI Fund keeps $GE as a whopping 27.12% of his portfolio. When you are managing other people’s money, keeping an allocation like that makes quite a statement. This is especially true given Hohn’s investing style, where he only chooses businesses that have multiple competitive advantages (multiple moats, not just one or two), and those which he is quite certain will be around 30 years from now. At 27.12% of his portfolio, Hohn’s certainty about this business is seemingly off the charts. Let’s see why. 🧵👇Image
2/10
$GE Aerospace operates in a duopoly, its counterpart being Pratt & Whitney. Lack of competition means that the business has plenty of room for continued growth and margin expansion. 3 out of 4 commercial flights are powered by GE’s engines.

The business model is quite simple. GE designs and sells jet engines for some of the biggest planes in the world (Airbus A320neo, Boeing 737 MAX, Boeing 777, and for defense/military fighters like the F-15, F/A-18, etc., bombers, and transport aircraft.) Once these engines are out in the world, GE has a lock on the high margin, recurring maintenance and services contracts needed for keeping them operational. Customers must use GE because these are some of the most advanced engines in the world, with proprietary technologies permeating the systems and GE being the only one equipped to properly manage their own technologies.Image
Image
3/10
What Makes the Business Special - Part 1

$GE does not make its money from selling its jet engines. In fact, it often has negative margins on each sale. But once the engines are in the planes, that is the gateway to the real profit-driving segment of the overall business– services. One of the things that makes the business special is the recurring, high margin service revenue that follows the sale of every engine put into play. Aftermarket services/spare parts accounted for 71.68% of revenue in Q3 2025, and flexed a 90%+ gross margin. Once the fleet is operating on these engines, they need maintenance, upkeep, and repairs for the entire life of the plane, which literally spans decades, from the people who designed and can understand the deep technology held within them. That’s the magic of the business model. Installing an engine is like signing a customer up for a long-term subscription that they have to keep paying, even if it gets more expensive. In fact, regulations mandate it. The real big profits kick in 7-10 years after an engine takes flight as service overhauls become mandatory. This brings us to our next point…Image
Read 10 tweets
Oct 13, 2025
Focus on what’s important and tune out the rest. These 5 charts tell the real story of $AMZN 👇🧵 Image
Over the past 2 decades, $AMZN has grown revenue at a 25% CAGR. Image
Over that time, operating cash flow margins have expanded, enabling the operating cash flow growth to outpace revenue growth. $AMZN Image
Read 7 tweets
Oct 7, 2025
[1/5]

$MELI A Peak into the Current Valuation 🧵
(Not financial advice)

Mercado Libre’s cash flows are key to understanding what the stock is worth. Specifically, it is important to try and wrap our heads around what the business could produce in cash right now if it suddenly decided that there was no higher priority than cash generation. We want to understand the underlying business’s true current ability to generate cash.

However, an assessment of the business’s potential cash flow generation is not exactly straight-forward. Mostly, it is complicated by the workings of its fintech arm, Mercado Pago, and capital investments into all facets of the overall business. Luckily, Mercado Libre outlined a succinct explanation of cash flows in the last quarterly earnings presentation which we can use to good effect. If we can go through the cash flows and look line-by-line through the lens of arriving at true potential cash flows, we can value the business. So, let's do exactly that. The image below outlines our logic for each line item.

Note: In the scenario where MELI is not expanding its loan book/credit portfolio, there is likely a portion of Restricted customer funds and regulatory requirements that would not be subtracted from Cash from Operations. Some of this value is likely customer balances, and some of it is likely regulatory reserves tied to the expanding loan portfolio. Ideally, we would refrain from subtracting out any portion of regulatory reserves that would have been tied to Loans receivable since this scenario sees no Loans receivable expansion. However, we can’t know the amounts here so we can’t change it. Just realize that the potential FCF number we come up with would likely be a bit higher for the quarter.Image
[2/5]

What did $MELI's business have the ability to generate this quarter if expansions were put on hold?

✅$2,917M operating cash flow
✅- $1,451M restricted funds
❌- $1,621M loans receivable
❌+$805M fintech funding
❌- $287M capex
✅+ $91M reconciliation adjustments
= $1,557M potential FCF for the quarter

Annualize the value:
$1,557 X 4 = $6,228M forward (potential) FCF

Currently there are 50.7M diluted shares outstanding
$6,228M / 50.7M = $122.84 FCF per share

The current $MELI share price at the time of writing is $2,148.75

$2,148.75 / $122.84 FCF per share = a forward (potential) P/FCF of 17.49

But is this cheap or expensive?

On the surface, this appears very cheap. The reason is that over the last 2 years, $MELI has grown its Cash from operations by a 36.59% CAGR. If we zoom out further, the CAGR increases. Mimicking a PEG ratio by substituting in FCF and FCF growth, the ratio looks to be < 0.5, implying extreme undervaluation.Image
[3/5]

However, when valuing a company, we have to be careful to always approach it from the perspective of alternative means of yield, especially those considered to have less risk, from the region that most revenue is earned. So, what am I talking about?

In Brazil, the economy where $MELI makes about 50% of its revenue, a 10 year government bond will yield a purchaser 13.89% today. For comparison, a purchase of MELI’s stock today at a forward (potential) P/FCF of 17.49 means a FCF yield of 5.7%. Even with robust growth, it would take years to push our FCF yield on cost up to a level of 13-14%.

Brazil’s 10 year bond yields have fluctuated over the years, and have actually ticked down recently as inflation eases. Over the last 15 years, the average yield has been around 11%, which is worth considering in a long-term compounding hold like $MELI.

However, when comparing the current FCF yield of 5.7% to the current, or even the average, Brazilian 10 year government bond rate, we can see that there is currently no margin of safety. This also implies that it would make perfect sense if $MELI’s market multiple traded lower, even considering its robust growth.Image
Read 5 tweets
Aug 26, 2025
$KSPI, Kaspi.kz AO, is a monopolistic, flywheel-driven e-commerce marketplace and fintech ecosystem in Kazakhstan that just keeps delivering results year after year after year, a true compounder. The stock is not without risks, but I think the risks are vastly outweighed by potential upside. I don’t hear anyone talking about this one. 👇🧵Image
The Kaspi.kz app has become an indispensable tool for daily life in Kazakhstan, integrating a vast array of financial, e-commerce, and government services into a single, user-friendly platform. Its ecosystem is built on three core business segments: Payments, Marketplace, and Fintech, which together offer a seamless and comprehensive user experience. As of the second quarter of 2025, the Kaspi.kz Super App has 14.2 million monthly active users in Kazakhstan. With the total population of the country being 20 million, this represents 71% of the entire country. That’s how dominant $KSPI is, and this figure continues to grow.

Although there is plenty of growth runway ahead in Kazakhstan for each of Kaspi’s segments, ambitions to grow the ecosystem led them to acquire 65% of Hepsiburada, an e-commerce platform in Turkiye. This effectively added about 400% to Kaspi’s TAM, and it only just happened a few months ago.Image
Each of $KSPI's segments are synergistic, strengthening their adjacent business segments in a flywheel that funnels revenue to the top line at every step. Everything feeds everything feeds everything feeds everything…
It's a self perpetuating machine. The better each segment does, the better the other segments do, and round and round it goes:

🔄As people make use of the app on a daily basis, merchants want to be where the people are, and people want to be where the merchants are. Having their payments and financial accounts on the Kaspi app makes transactions seamless for both patrons and sellers. People buy and the number of merchants grows. Sellers utilize fintech solutions like loans to bolster their businesses. They advertise through the Kaspi app and people buy more, which means more total payments volume for the payments segment. Both customer and seller satisfaction is high, and so the flywheel goes round and round.🔄Image
Read 15 tweets
Jun 21, 2025
$AMZN is one of the greatest businesses on Earth. Surely, with a big wide moat, global expansion, and increasing margins, there is a lot to look forward to. But is it undervalued? The usual look into P/OCF doesn't even begin to tell the story. The thread below outlines a different take on the valuation which suggests the stock will casually triple over the next 5 years. 👇🧵
Chart from @stock_unlockImage
While most analysts are hyper-focused on the extremely impressive AWS business, zooming in on the e-commerce business shows something surprising. Market research suggests that the US e-commerce TAM will expand over the next 5 years to $1.8T. What does this mean for $AMZN?

Source: statista.com/statistics/272…Image
The strength of $AMZN's delivery network, subscription service, and seemingly unending selection means that their ownership of market share has been steadily increasing by anywhere from 0.1% to 1.35% each year. We are going to do a little math soon, so for the sake of conservatism we will use a 0.3% growth rate for market share.

Source: emarketer.com/content/amazon…Image
Read 8 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(