The Bird is Chirping and the Bulls are Tweeting
Why Twitter’s new ‘Super Follow’ feature could generate over $2 billion in annual revenue by 2025
Disclaimers — Not investment advice. Views are solely my own. I am a Twitter shareholder.
$TWTR has been blowing up on my Twitter. The company announced at its recent investor day that it would be adding a Super Follow feature some time in 2021, which will allow users to charge a subscription fee for exclusive content (e.g., exclusive tweets, Revue newsletters, Spaces hangouts, etc.). The proposal has been watering the mouths of social media influencers and Twitter investors alike, and helped the company’s stock continue a historic run that began back in January.
Notwithstanding a dip yesterday, Twitter shares are up over 35% YTD and over 110% in the past 12 months (as of this writing). Importantly, it also finally eclipsed its all-time high (previously set back in December 2013 (!) a few weeks after its IPO) and the narrative around the company seems to be shifting after a long period of failed innovation and under-monetization.
At investor day, CEO Jack Dorsey hinted at lots of groundwork (including an ‘ad server’ infrastructure rebuild) to get the company to this point, where it now seems poised and committed to driving innovation and accelerating product-led growth. With those pieces in place, the company is targeting a trio of ambitious long-term goals including doubling development velocity (as measured by features shipped per employee that drive user or revenue growth), hitting 315 million monetizable daily active users (mDAUs) by Q4 2023, and at least doubling annual revenue to $7.5 billion by 2023.
Matthew Ball’s tweetstorm on cultural determinism articulated well the importance of this fundamental shift of focus to product development and innovation, especially as it relates to premium features that will enable Twitter to capture a greater share of the enormous value that it creates.
The business case for ‘Twitter Premium’
Investors and users alike have been openly wondering why Twitter hadn’t added a premium ad-free version years ago — what I’ll refer to throughout as ‘Twitter Premium’. When you consider the platform’s large and engaged user base, it doesn’t take much imagination to see a path to significantly increasing revenue with what seem like fairly conservative assumptions.
Drop in whatever placeholders feel right, but the general consensus appears to be that there is a secular shift going on in media with revenues shifting from advertising to paid subscriptions. I suspect this will only accelerate as consumers become increasingly aware of — and willing to pay for — data privacy.
The long-time critique — that seems to be reversing in the current narrative — is that Twitter’s lack of product innovation has kept it from fully realizing the monetization potential of its platform in the same way that its peers (especially Facebook and Google) have been able to. When you consider average revenue per user (ARPU), Twitter lags behind social media and entertainment comparables, despite evidence of massive mindshare and engagement.
So the billion dollar question is — how many people would subscribe to Twitter Premium and how much would they pay? Personally, I don’t seem to notice (or be annoyed by) ads on Twitter to the same extent that I do on Instagram, Reddit, and (especially) YouTube. Sample size of one, but it leads me to wonder if an ad-free offering could achieve the same level of adoption as other platforms, such as Spotify, for example, where 45% of monthly active users are premium subscribers. On the other bookend, the New York Times reported 7.5 million paid subscribers in 2020, which would represent about 5% of their monthly visitors (per similarweb).
Interestingly, Pew Research Center studies have found that Twitter users tend to be more affluent than the general population. They also found that the top 10% of users generate 80% of all tweets. Together, these findings suggest to me that at least 10% of daily active users (that’s daily, not monthly) paying $5 or more per month for a premium offering is probably very achievable.
Thus far I’ve been focusing on ‘Twitter Premium’ as a site-wide, ad-free offering. Other versions of a paid offering include adding a paywall to trending hashtags or providing real-time access to premium users and delayed feeds for everyone else (a la stock market data feeds). While these shift more value behind the paywall, the whole is arguably less than the sum of the parts. As highlighted by Jack Dorsey and other senior leaders in the company’s recent investor day, Twitter’s purpose is to serve the public conversation. The objective is to increase participation in social discourse and make it inclusive of as many voices as possible. The company’s initiatives around user trust and safety, including Bluesky, reflect this as a core value of the company. And I tend to agree with them — while adding site-wide paywalls such as these could be extremely lucrative for the company, it would also detract from the fundamental value of the platform. Twitter is what’s happening right now, and that inherently relies on real-time posts from a diverse base of engaged users.
For better or worse from a shareholder perspective, Twitter has created a public good that should not be restricted to those who are willing and able to pay for access. That being said, the broader debate around whether and how social media platforms should monetize their users (incl. through collection of personal data) is an important and complex one, and outside the scope of both this post and my expertise…
Enter Super Follows. As compared to a site-wide paid model, the Super Follow model bifurcates public and private conversations. This important nuance will enable (and incent) creators to build highly engaged private communities of aligned individuals, while maintaining the integrity of the public conversation. Social discourse and breaking news should be equal access. Niche, interest-based communities can be pay-to-play. “Choice” was one of the callouts in Twitter’s investor day discussion on trust, and I think that is exactly what Super Follow provides.
“Would you Super Follow me?”
Along with these philosophical arguments, I think Super Follow has much more potential than other iterations of Twitter Premium for three key reasons:
- Community-led demand generation and engagement
As compared to selling paid subscriptions directly, on which Spotify, for example, spent $1.2 billion (13% of sales) in 2020, Twitter is effectively outsourcing and distributing sales and marketing to its creators. In doing so, it leverages its strongest competitive advantage — the incredible breadth and depth of its network and communities where users are highly engaged on subjects that they care about. Each of its hundreds of thousands of creators becomes a partner with vested interest, bringing highly targeted content to monetize their following, and by extension, the Twitter platform. This feels a lot more compelling than removing ads from my feed and for that reason I believe Super Follows can achieve much higher adoption than other media paywall analogs.
2. Dynamic pricing
Whereas a site-wide premium subscription applies a flat-rate to all users – say $5 or $10 per month — the customizable pricing and a la carte nature of Super Follows means that it will more fully capture each individual’s propensity to spend on creator content. As seen on other digital content platforms (notably gaming), the top 1% of users (whales) generate more revenue than the bottom 20 / 50 / 80% combined. On the other end of the spectrum, this model also has the potential to pick up the long tail of people that are only willing to spend $2 or $3 per month. Even with a large percentage of paid users spending less than the alternative flat rate, the whales pull up the average spend per paid user and the long tail captures incremental users. I believe this effect may be even more pronounced on Twitter since exclusivity and social signaling are Veblen goods — there will be creators that charge $999 per month because they can (assuming Twitter doesn’t implement policies against this) and there will be users that subscribe to them because they can and they want others to know that they can.
3. Net new monetization surface area
Another important benefit of the Super Follow model is that it is predicated on adding new products, features, and experiences, as compared to other premium models that are based on taking something away — whether that be ads or existing access. Not only does this mitigate the risk of user backlash and churn, but it also means that subscription revenue does not come at the expense of ad revenue. In the ‘Twitter Premium’ model, each user that converts to paid is generating higher ARPU, but is also removed from the pool of mDAUs that drive ad revenue. I would hypothesize that the users converting to paid are also among the higher value advertising targets as well, so this impact could be disproportionately large (e.g., a 10% paid adoption rate results in more than a 10% reduction in ad revenue). In this way, paid subscriptions cannibalize ad revenue, whereas Super Follow revenue is fully incremental.
For these reasons, I believe Super Follows (and tipping) have the potential to achieve both higher adoption and higher paid ARPU than Twitter Premium. With all that being said, there is one critical (economic) drawback to Super Follows that I haven’t touched on yet… take rates.
Super Follow is at its core a revenue share. For each dollar of Super Follow subscription revenue, some percentage is first paid to the creators that generated that revenue. The amount that remains after paying this cut can be thought of as Twitter’s ‘take rate’ akin to other digital marketplaces.
Similarly to paid adoption rates, take rates / revenue shares vary widely across different industries and platforms. Whereas a blanket paywall subscription would imply a take rate of 100%, take rates of Super Follow analogs range from from as low as 5% or 10% on Revue and Substack, to 50%+ on Twitch and Roblox, for example.
Equity research analysts seem to be expecting Twitter to land on the lower end of the spectrum due to competitive pressures (and anchoring on Revue and Substack). I would argue Twitter provides a much stronger platform, audience, and product to creators than these tools on a standalone basis. Furthermore, as Ben Thompson (Stratechery) has written about extensively in his posts on aggregation theory, platforms such as Twitter that have aggregated the demand-side (consumers) stand to capture substantial economic profit from the rest of the value chain because they have more bargaining power. While there are several point solutions for creators to monetize their following (e.g., Substack, Patreon, Cameo, Etsy), I would argue there are few that are able to provide the full product suite that Twitter is building (i.e., tweets, Analytics, Revue, Spaces) and fewer that have the same immediately monetizable captive audience.
Can $TWTR keep flying higher?
As mentioned at the outset, the stock has been on a tear over the past year and especially in the last 60 days. This inevitably leads to the argument that the upside of Super Follows is “already priced in.” Let’s investigate that further…
Above is a simple forecast that adds Super Follow (and tipping) revenue to the existing consensus estimate (pre-Investor Day upgrades [1]) by applying assumptions around mDAU growth, Super Follow adoption rate, average spend, and take rate. The intent here is less about picking the right numbers and more about getting a sense of magnitude of upside if we’re in the right ballpark (“directionally right, precisely wrong”); see Notes [2] if interested in sensitivity tables. Even with only four basic assumptions there is a lot of variability, but the key takeaway for me remains that Super Follows represent a substantial incremental growth opportunity. As mentioned earlier, the important nuance in this scenario is that there is no cannibalization of ad revenue; these revenues are 100% incremental. Under the assumptions outlined above, Super Follows would generate over $2 billion in net revenue by 2025 and represent over 20% of Twitter’s total sales.
How does this translate to share price? With some more simplifying assumptions (i.e., same net debt and shares outstanding as today) it would appear that a 20%+ return is achievable IF Twitter continues to trade at a similar multiple to today. That being said, as you can see from the sensitivity table on the right, outcomes vary widely depending on the prevailing valuation multiple. And, when you compare TWTR’s current multiple to its trading history, the prerequisite that current multiples hold up might be a big if.
Over the past five years, TWTR has basically never traded at its current valuation multiples (see Notes [3] for valuation charts). I know, stonks only go up, but if there is a reversion to the mean in terms of how the market values Twitter on a revenue multiple basis, then now might not be a great time to invest. That being said, if the company can deliver on its goals, demonstrating community-led growth and a more durable revenue model, then historical multiples may not be totally relevant and you could very well see a new anchor being set. By comparison, SNAP currently trades at ~25x ’21 revenue and PINS at ~19x ’21 revenue, though I’m not suggesting these are all that reasonable either…
Share prices aside, the fact remains that Twitter has built a platform that creates real value for over 200 million people every day. If Super Follow and other product bets are successful, the coming months and years could mark an inflection point in the company’s history. Whether you are passionate about community, product, marketing, growth, ecommerce, investing, or something else, there will be tons of learnings coming out of this experiment, so I’m grabbing a pen and taking some notes!
Please do @ me on Twitter if you liked this post (and if you didn’t like it for that matter, even better!)
Notes
[1] By layering incremental Super Follow revenue on top of existing consensus estimates, the scenario outlined above assumes that consensus estimates prior to Investor Day did not include paid subscription revenues and were predicated only on advertising revenue. While all of the equity research reports that I have read include only advertising and data licensing revenue, I concede this may be an aggressive assumption. One could argue that it was only a matter of time until Twitter adopted some sort of paid subscription model and that this shift has been contemplated for some time now, implicitly baked into longer term growth rates.
[2] Super Follow forecast sensitivity tables based on different assumptions on adoption, average spend, and take rate
[3] Twitter valuation as compared to peers (SNAP, PINS, FB, GOOGL) and its 5 year trading history
