Description

Twitter went public in November 2013 at $26. Its first trade was at $45 and closed the year at $60. Eight years later, the stock is at $36. Not split-adjusted. No dividends were paid. Just a stagnant, money-losing performance from a company that, in 2013, was theoretically perfectly positioned for what would become the greatest value creation tailwind in world history. 

Twitter’s public shareholders captured none of it - clown car in a gold mine.

What follows is a discussion of why that may be about to change.

Thesis:

Note: Twitter just reported Q4 numbers, basically in line with guidance. It maintained its high-level guide through 2023 and announced a $4B buyback (market cap is $29B). It plans to accelerate $2B of the buyback and do the balance patiently. My write-up discusses many things through Q3, since I wrote most of it pre-earnings, but nothing is materially different. Do your own due diligence. Twitter has been a clown car and its investors prefer losing to winning. Employees dilute your opportunity by receiving tons of compensation in the form of your stock. I would not become a shareholder if I were you. Caveat emptor.

Happy Valentine’s Day…I give you carbon. But is it coal or is it diamond? Probably coal.

Backdrop:

Twitter has stumbled through many real challenges: mired by inconsistent leadership, poor underlying tech, a politicized platform, and missed opportunity, one could argue it has failed to live up to its promise.

That said, it may surprise some to learn that from its IPO year of 2013 through YE2021, Twitter has 7.7x’d revenue, posting a 29% revenue CAGR during that time…and still growing. 

Twitter just reported 2021 revenue of nearly $5.1 billion, 37% higher than the negatively impacted COVID year of 2020 (which was itself +7% over 2019). The company forecasts 20%+ growth over the coming few years, including $1+ billion of topline growth in 2022 (“low- to mid-20s” adjusted 2022 revenue growth - adjusted for Twitter’s sale of MoPub in early January 2022). Its 2023 revenue goal of $7.5B remains unchanged, implying growth accelerating to ~$1.5B in 2023 (high 20s growth).

The biggest challenge Twitter’s stock has faced is not execution. It’s not poor management. It’s not platform health. It’s not ridiculous SBC. While each of those is a very real issue, the biggest challenge was an insanely priced stock in 2013. The first trade post-IPO was 40x 2013 sales and it ended that year at 60x 2013 sales. SaaS owners may want to take note.

Today, at $36, it is 5.5x TTM adjusted revenue (EV of $27 billion). Twitter’s 2023 revenue goal is $7.5B - you can do your own math. If Twitter executes, it can be a multi-bagger over the next several years and compound from there.

The Twitter of 2013 had a nosebleed multiple and problematic tech.

The Twitter of 2022 has an undemanding multiple and an overhauled, increasingly modern tech stack. It has the opportunity to grow for years and years.

Assuming continued growth and reasonable estimates of potential margins, I believe the stock could be a 3x in the next four to five years.

Brief History:

Twitter was founded in 2006 by Jack Dorsey, Noah Glass, Evan Williams, and Biz Stone.

Jack was Twitter’s first CEO. He was soon deposed by Ev Williams, who took over the reins. In 2010, Ev was usurped by disgruntled VCs who appointed Ev’s friend, and Twitter COO, Dick Costolo as CEO. Its rapid early growth masked problems beneath the surface: it was accumulating massive tech debt while simultaneously struggling to understand its market position and define its moat. Despite this, Costolo would oversee Twitter’s November 2013 IPO - it was the sexy SaaS of its day.

Inevitably, Twitter’s problems would surface and, in 2015, Costolo was ousted and Jack triumphantly returned - first as “interim,” then permanent, CEO. Jack was dual CEO of Twitter and Square/Block (which he also founded) until four months ago. Apparently burned out, he resigned in November and was replaced by his hand-picked successor, CTO Parag Agrawal. Twitter simultaneously named Salesforce’s newly minted Co-CEO - the highly respected Bret Taylor - as its new Chairman.

In 2015, Jack took over a total shitshow - Twitter lacked focus, was losing users, had massive platform health problems (e.g., women and minorities being harassed), terrible PR, and disastrous tech. He was given a bad hand, offset only by Twitter’s cultural relevance and passionate user-base.

Jack would immediately focus on simplification. This meant eliminating non-core assets (e.g., Vine was deprecated), laying off 9% of its workforce, launching an algorithmic timeline that gave it more control over user experience, and increasing the character limit on Tweets to 280 to allow more expressiveness.

By late 2016, Jack made the decision to rebuild much of the tech stack. For the next four years, Twitter’s “growth” spend was really on going back in time to fix its platform - this meant rebuilding Twitter’s code from scratch, in a modern, agile system, while simultaneously maintaining (and improving) the legacy tech so that Twitter could continue operating in parallel. The bulk of this took through 2020 (with the “ad server” modularization being the capstone project). Though certain elements remain on old tech (e.g., DMs), much of the heavy lifting is now behind it.

During this process of “going backward” to repair itself, new feature development was largely sidelined as rebuilding existing functionality took priority. As platform health improved, Twitter’s user losses began to reverse, growing nicely in 2019 and then slingshotting through the pandemic. 

Twitter had to go backward to go forward - by mid-2020, it caught up with the times. Since then, it’s been full speed ahead. It has put out more new product features in the past year and a half than in its entire history. 

While Jack took over a terrible hand in 2016, in 2021 he handed Parag strong cards and a rebuilt stack. [see what I did there!]

2023 Goals:

Given its comparatively small scale and history of poor execution, there are two major questions:

1. What is the long-term growth opportunity?

2. How much of that growth will accrue to owners, given low margins and high SBC?

Regarding A), the overall $600B global digital ad market (ex-China) is expected to grow teens and Twitter is taking share within that, as its existing brand advertising strength combines with improving targeting, measurement, and down-funnel performance ad products. Likewise, Twitter is launching a series of subscription products that have the potential to generate hundreds of millions of dollars of very high incremental margin revenue.

At its Analyst Day in early 2021, Twitter laid out three overarching medium-term KPIs:

1. End 2023 with 315 million mDAUs (vs. 192 million at YE 2020 or +64% 3-year growth)

2. Generate $7.5 billion in 2023 revenue (up >100% from 2019’s $3.7 billion)

3. By 2023, double development velocity as measured by successful features delivered per employee that impact mDAU or Revenue

That leads to B): how much of that growth will accrue to owners? For years, Twitter has given a hand-wavy long-term margin guide of 40-45% Adjusted EBITDA (adding back SBC). It reiterated that range during the Q4 earnings call. Unfortunately, the company provides no context, no bridge, and no timeline. This lack of color is dumbfounding and margins are set to worsen in 2022 as it invests aggressively in hitting its 2023 goals. Skepticism is warranted.

The balance of this write-up discusses the user, ARPU, and margin opportunities.

Usage / mDAU:

Twitter is in the midst of a transformation from a microblogging service to a Creator business platform. In many ways, it already was this, without formally acknowledging it. 

This shift means Twitter’s most distinctive role is helping users find the people and communities they care about. Creators are a magnet for Users and Users are a magnet for Creators: Twitter’s job is to help Users find Creators, serving as a User acquisition vehicle for Creators. When done well, this happens on the platform and builds lock-in for both parties, as the challenge of recreating that connection elsewhere is not worth the switch.

Creators benefit as their influence grows. Users benefit as their connections and knowledge deepen and grow. The resultant network effect is Twitter’s core moat. Nearly everything Twitter does should enhance this self-reinforcing loop. Twitter is increasingly focused on helping Users discover content they love and value while helping Creators produce, organize, and monetize that amazing content. Neither party should ever have a reason to go elsewhere to enhance their connection.

To drive this, Twitter is increasingly taking control of the user experience, the core of which is the Twitter Timeline (your primary feed). 

Historically, your Timeline was a chronological feed of Tweets by people you follow. This harkened back to Twitter’s founding as a subscription to public text messages. But, as the service scaled, that chronological experience became a firehose of disconnected announcements. By using an algorithmic timeline, Twitter can combine the best of follow-based relevance with Twitter’s understanding of what is valuable to you, right now.

The launch of “Topics” two years ago ushered in a new era of Twitter’s ability to personalize. If you love the Brooklyn Nets, rather than having to piece together the key Creators (fans, players, beat writers, etc.) who tweet about the Nets, you can now follow a Topic called “Brooklyn Nets.” You could also zoom out to Topics called “NBA,” “Basketball,” “Sports,” “Brooklyn,” or “New York.” Each of these provides strong signal to Twitter about your interests, both for engagement and monetization.

Twitter has a stated goal of “hundreds of thousands” of Topics. Today, there are over 14,000 in eleven languages and more than 280 million accounts follow at least one Topic (up from 230 million in Q3!). Each new Topic, and your related interactions, brings Twitter a deeper machine-learning-powered understanding of who you are and what you care about (likewise, its understanding of others with similar interests). This enables continuous improvements in personalization and targeting.

Launched just two years ago, Topics was one of the first major features out of the tech stack rebuild. Since then, we have seen accelerated development velocity. In the past 18 months, Twitter has launched more features and enhancements than in its entire history. Those include:

• Fleets

• Tipping

• Birdwatch

• Communities

• Super Follows

• Live Shopping

• Recorded audio

• Subscription (Blue)

• Professional Profiles

• User-controlled safety tools

• NFT profiles and wallet linking

• Edge-to-edge visual mobile app

• Google and Apple ID SSO Login

• Improved photo and video sharing

• Simplified, interest-based onboarding flow

• Shopping module for professional accounts

• Spaces (live audio rooms) and Ticketed Spaces

Most of those remain in early beta. Some will be killed (e.g., Fleets were launched and shut down) while others may become key features. Twitter will spend 2022 iterating within those, improving many features to a more functional level while continuing to launch new experiments. Its improved development speed is obvious to anyone watching.

Twitter committed a few years ago to “build in public.” This means that most features launch early and small, with Twitter intending to watch how they are used, improving them to meet real-world needs. Human nature being what it is, this actually creates new user frustrations, as people gnash teeth over a new product’s time in beta purgatory (“Super Follows launched in September and still haven’t integrated Private Spaces - wtf?!”).

2022 should provide several prominent product overhauls and new features:

• DMs

• Private Tweets (“Flock”)

• Tweetdeck: expect it to become a premium subscription product

• Newsletter and Long-form (Revue: Twitter’s substack competitor; and “Articles”: a new >280 character on-platform product that is being developed)

• Search and Discovery: in particular, I expect we will see more full-screen video discovery (video is engaging and ad-friendly - i.e., TikTok/Reels)

Each of these is a substantial engagement opportunity. In an effort to accelerate mDAU growth, Twitter is pairing its growing engagement with improved onboarding flow (removing friction to customizing a valuable Timeline) and a firmer nudge to register (e.g., limiting logged-out user access).

Twitter is innovating faster than at any point in its history.

ARPU / Digital Ad Market Opportunity:

In addition to the user-facing experience, Twitter is rapidly improving its ad tech. Behind this was a complete rebuild of its ad server in the 2019-2020 timeframe. That rebuild is allowing faster development velocity and more experimentation. The improvements include the beginnings of down-funnel performance ad products designed to unlock the massive SMB opportunity it has historically struggled to touch. 

Twitter today has ~5% share of US digital ad spend by Large brands, but <0.5% share in SMB. SMB represents more than half of the digital ad TAM. The result of Twitter’s imbalance toward Large is that “less than 15%” of Twitter’s ad revenue is from SMB. Compare to Meta which is 75% SMB.

Today, 1/3 of the ad budget of a typical Large business is digital, while digital is 3/4 for SMBs. Twitter has millions of SMBs on-platform, but generates little revenue from them: its products have historically failed to meet SMB needs.

SMBs need four things: self-serve, narrow-targeting, DR/e-commerce, and measurement. Twitter is investing in each of these in order to unlock a 10x opportunity in SMB. Its stated long-term aspirations are to achieve a 50:50 Large:SMB revenue mix. 

The performance ad market, which includes much of display and elements of search is the fastest-growing element of the global ad market. Performance ads are theoretically only bounded by the contribution profits of the companies that use them to drive sales. As performance ad technology has improved, it has continuously expanded demand and likely will for years and years to come.

Digital Ad Market:

Global ex-China digital ad market is expected to CAGR teens through 2025, reaching $500 billion that year. Direct Response (DR) and performance advertising is growing within that and could approach $350 billion. 

Framing Twitter’s opportunity another way:

See Note on ARPU

In the US, Twitter has a more affluent average demographic than Meta. However, as you can see above, it earns 1/4th the US ARPU of Meta. Twitter also has 1/5 the user base. Those two factors largely explain why Twitter is 5% as large as Meta.

Volume x Price:

The upside of being smaller is that Twitter has a long runway for growth if it executes. Today, Twitter is a share-taker in a growing market.

Volume: Unlike Meta, which has a fully penetrated (i.e., stagnant) user growth profile, Twitter is growing users in the double digits. It ended Q3 2021 with 211 million mDAUs and Q4 with 217 million. Despite extreme skepticism by Wall Street, the company continues to affirm its 315 million mDAU goal for the end of 2023, which implies mDAU growth accelerates significantly. I am likewise skeptical of the 315 million target, but even coming well short could still represent substantial user growth.

For 2022, Twitter has indicated it expects mDAU growth to accelerate through the year. When pressed, new CEO, Parag Agrawal, said twice on the recent earnings call that Twitter has “line of sight” on the 315 million target, based on its assessment of the top of the funnel and improving conversion techniques and trends. We shall see.

Price: Improving ad tech and new subscription businesses are powering a secular ARPU opportunity. Twitter should be able to grow ARPU through nearly any cycle, as it begins to unlock the SMB opportunity. SMB is a self-help execution opportunity: a 10x opportunity by simply catching SMB up to its Large revenue. In the meantime, Large will not be standing still as Twitter grows users and improves targeting and measurement.

This combination of Volume (user growth) x Price (ARPU) provides the ingredients for a robust growth algorithm for years to come.

Potential Margins:

But to what end? 

Twitter’s operating margins (EBIT) have been in the low double digits and are expected to compress in 2022 as it invests aggressively to hit its growth targets: if we are playing for 10% EBIT margins, it is not a game worth playing.

Twitter’s challenge today is the need to balance the spending required to increase usage and ARPU against the fact its shareholder base is rightly critical of its history of profligate spending (both cash costs and SBC).

What we know for certain is that Twitter is investing aggressively to grow. Disentangling what portion of its OpEx is Grow vs. that which Maintains is impossible - our best hope is to estimate. I believe Twitter is wasteful and inefficient, but for this purpose, I’m assuming the current OpEx won’t become more efficient. Rather, that OpEx is comprised of two elements: spend to Maintain and spend to Grow. In essence, there is a smaller amount Twitter could invest if it were to choose to simply Maintain. Said another way, we know Twitter is spending aggressively to Grow. We just don't know how much.

We can frame Twitter’s Gross Profit as the pool of income available to Maintain and Grow. If we benchmark Twitter against Meta, we see that Twitter has 15 point lower gross margins (65% vs 80%) - this is driven by Twitter’s higher content partner revenue shares and content spend. 

In 2021, Twitter spent ~$3 billion on OpEx. $3.0 billion is ~59% of $5.1 billion of revenue. Given a 65% Gross Margin, that leaves a 6% EBIT margin for owners. But the 59% is a combination of Maintain + Grow - a universe of possibilities (and narratives) exists within it.

Compare Twitter to Meta's 2021 "Family of Apps" income statement:

• $116 billion - Revenue for FoA

• $57 billion - EBIT (49% EBIT margin) for FoA

• OpEx was 32% of FoA Revenue (assuming 81% gross margins)

That 32% OpEx spend actually supports aggressive investment in growing Meta's Family of Apps businesses (FoA revenue grew at a 28% two-year CAGR through 2021).

To me, Meta's 32% total FoA OpEx spend tells us something about what is possible at Twitter - Twitter is spending 59% to Maintain and Grow. If Meta can grow rapidly with a 32% OpEx spend, I expect Twitter could spend 32% and at least Maintain.

If so, Twitter is spending at least 27% on growth OpEx. That also implies it could earn well over 30% EBIT margins in a low-/no-growth environment (likely higher, as Twitter clearly does not run with a lean, efficient mindset).

Said another way, I believe Twitter can earn 35%+ EBIT margins in a "Maintain" environment. (Note: 35% EBIT comports with management’s color on 45% long-term Adjusted EBITDA margins). 

If Twitter earns 35%+ contribution margins on revenue growth, then each $1 billion of revenue growth contributes at least $350 million of incremental normalized EBIT. 

If we grant that Twitter spent 27% on Grow in 2021, that cost $1.37 billion ($5.08B x 27% = $1.37B). The $1.37 billion Grow spend earned Twitter an incremental $1 billion of revenue and $350mm of Maintain EBIT: a <4-year payback. SaaS-like! This is the type of math that has made Meta a monster.

On Twitter’s $1.1 billion in estimated 2022 revenue growth, that’s a further $385 million in contribution margin. 

And again in 2023. And 2024. And so on.

If a growing, recurring EBIT trades at 15-20x, Twitter is adding $5.5 to $7.5 billion of compounding value each year on an EV of $27 billion. 

This implies a potential 20-25% long-term CAGR - a game worth playing.

The below graphic assumes Maintain margin is 35% and shows what spend on Grow has been since 2018 (blue line, right hand axis). The orange line (left hand scale) shows return from incremental Maintain EBIT generated (in the same year) at 35% margins, assuming 15x EBIT. The grey line shows the same, but shows the trailing two-year MOIC (to smooth out the Covid impact).

My Simple Mind - Alternative Ways to Look at Valuation:

1. Twitter captures its opportunity - By the end of the decade, Twitter reaches $15 billion of ad revenue (implying low teens CAGR from 2023). At 30% EBIT margins, while continuing to invest in its long runway for ongoing growth, Twitter’s ad business could trade at 20x EBIT ($100B of value). Likewise, Twitter has a large subscription and data opportunity in front of it that could be worth $20+ billion by the end of the decade. Along the way, it should be able to generate $10+ billion of cash (+$2B of starting net cash). Discounting that $132B of EV back at a 10% rate is $112/share to year-end 2025 (close to a 3x). 

1. Framed differently, a $100 stock in 4 years (YE 2025) implies Twitter trades ~6.7x 2026 revenue, well below its average EV/Revenue multiple since 2018. 

2. Flat Multiple: TWTR returns approximately its revenue growth rate: 20%+ CAGR. 

3. No Return / Flat Stock: A flat price of $36 at YE 2025 would represent an EV/Revenue multiple of <3x 2026. That would be below the lowest EV/Sales multiple Twitter has EVER traded at (3.3x - when it was posting negative topline growth in 2016-17).

4. Lose Money: Something worse than #3 happens. Growth stalls out, multiple compression comes once-and-for-all as Mr. Market loses all hope in the business.

In a downside scenario, Twitter remains an incredibly desirable acquisition target. Its user scale, premium demographic, high engagement, and cultural importance are massive assets. While Meta, Alphabet, and ByteDance almost certainly cannot acquire it, Salesforce, Spotify, Snap, Microsoft, Disney, and PE potentially can. At a minimum, I believe its current value is easily underwritten by acquirers. 

Of course, TWTR stock can trade down, but I believe it ultimately has a put option as the lowest valued scaled, highly strategic consumer internet asset in the world. With Elliott and Silverlake pushing on the Board, I expect if current management doesn’t successfully guide the business, new ownership will. I do not want this, but it’s a possibility.

Caveat Emptor:

I want to be very clear: I do not expect management to focus on margins in the near term. Management is “all-in” on the 2023 KPI goal. If Twitter trends positively toward that goal, I suspect Mr. Market will reward it and the company will continue to invest heavily in growth.

For me, the key insight is not to understand the precise proportions of Grow vs. Maintain, but to acknowledge some meaningful portion of OpEx is clearly Grow. While we don’t know what medium-term margins will be, we know Twitter is a) already profitable; b) investing heavily in growth; and c) achieving growth. That allows us to estimate what margins can be:

• If you think Maintain margins are lower, that implies the amount spent today on Grow is also lower (implying efficient growth);

• If Maintain margins are higher, the proportion of today’s OpEx spent to Grow is also higher, implying Twitter is spending more to gain even more meaningful future EBIT.

If my napkin math of potential margins is in the ballpark, Twitter is spending $4 today to gain $3 of revenue tomorrow (and $1 of Maintain EBIT): a quality investment. That $1 of EBIT is probably worth $15 to $20. A 3x to 5x ROI.

Even if Maintain margins are half what I estimate, it’s still a reasonable payback (spend $4 to generate 50c of EBIT, worth $7 to $10).

At 15x EBIT, Maintain margins that simply offset the spend would be 9% margins. Above that is a positive NPV. It’s nearly certain that Maintain margins are much higher than 9%.

One last point - Management Incentives: management bonus compensation includes weightings for revenue growth, mDAU, stock performance, and GAAP EBIT margin (fully loaded for SBC). 

Risks and Frustrations:

1. Clown Car: Twitter’s history is one of letting owners and users down over and over again.

2. Twitter is Wasteful: It spends too much on SBC and has a headcount that leads to some of the worst per capita efficiency metrics in the industry.

3. SBC: I already said this, but I will say it again - Twitter is giving employees 2.5% of the business in shares every year. While below Snap’s even more egregious number, mature operators like Meta are closer to 1.5%.

4. 2023 KPI Miss:

1. While the $7.5 billion of revenue seems within reach, the 315 million mDAU goal is clearly a stretch. No coverage analyst believes it will hit 315 (and most assume revenue falls short - consensus is $7.3B). The 98 million incremental mDAUs needed is ~12.5 million per quarter for the next eight quarters. Twitter has only twice achieved that level and both were Covid beneficiaries (Q1 and Q2 2020). I do not expect the company to hit 315 million by year-end 2023. 

5. Take-Under: Elliott and Silverlake are on the board. If Twitter hits an air pocket, we risk having the business taken away from us by someone prepared to run it more efficiently.

6. New CEO: Parag has no experience running a large team, much less a large public company.

7. SBC: It really is egregious.

Notes:

On ARPU: 

Twitter and Meta report different quarterly ARPUs - Twitter does it based on mDAUs whereas Meta uses combined Revenue (across IG, FB Blue, etc.) divided by Blue MAUs. They offer other variations, but don’t give geographic detail, so I am adjusting that “divided by Blue users.” (Meta uses UCAN whereas Twitter uses US.) There are puts and takes to making adjustments, but - for comparability - the above compares US for Twitter to UCAN for Meta and uses Meta’s total revenue divided by DAUs (rather than MAUs). Keep in mind mDAUs < DAUs for Twitter, which slightly inflates Twitter’s ARPU vs. Meta. OTOH, I’m only using ad revenue for Twitter, whereas Meta uses total revenue across all platforms and revenue types.

Additional Notes:
I wrote pages and pages that I ended up cutting, since I think most people already have a working understanding of Twitter and prefer concision. If you’d like more depth, feel free to ask in the comments or search me out on Twitter. Likewise, I did a podcast that covers some of these highlights. You can find that here.

Twitter stock is a killing field - do your own due diligence. 

See Cuyler1903’s great April 2020 VIC write-up and 430+ reply conversation (

Catalyst

• 20%+ growth persists
• Twitter mDAU growth inflects
• Twitter ARPU inflects
• Twitter discusses a path to its long-term margin profile
• Acquisition