Palantir: The Most Mysterious IPO of 2020

Preface: bit of a lighter, less analytical one this week, as Fanny I take a little time away from the office. Back to regularly scheduled programming next Sunday, though. Enjoy!

Palantir is one of those companies that almost everyone (in the tech world, at least) has heard of, but very few actually know much about.

This blind spot when it comes to the secretive deep-tech company began to come into the spotlight through the first half of 2020, as there have been talks of a possible Palantir IPO on the horizon.

Sure enough, after years of quiet rumors about the company one day going public, and hopefully offering even just a glimpse at the intriguing inner workings of the firm, Palantir quietly filed documents with the SEC with intent to go public later this year. No date is set as of yet, and no public S-1 has come out of it so far, but the news has definitely pushed the company into the tech and investing world spotlights.

This week, I want to dive into a look into Palantir’s story, what they actually…do, and explore some recently leaked figures from the company ahead of its public debut. With that said, here we go!

Some numbers

TechCrunch recently published an article with some key figures from Palantir’s upcoming S-1, “leaked” (quotation marks very much mine) by a company insider. The first look into precise numbers from the company was long-awaited and, honestly, kinda surprising!

Sure enough, the leak offered an unexpected key figure, in the form of a net loss of roughly $580M in 2019 on top-line revenue upwards of $742M for the period. This number was maybe the most surprising from this most recent leak. Tech companies incurring significant operating losses is nothing new, of course. What is new though, is a tech company running a 78% loss going into its 17th (!) year of activity.

In fairness, revenue is trending upwards, up over 25% YoY from 2018, and net losses are reducing, in turn, gradually — from a 97% loss in 2018.

Equally surprising was the discovery that around $900M of the combined losses spanning 2018 and 2019, were attributed to spend on sales and marketing. For a company known for having primarily governments as its core clients, the figure raised some eyebrows. Of course, there’s no breakdown available quite yet of how this spend was calculated. Ex-insiders at the company claim that a large part of this cash burn can be attributed to the extensive technical development carried out in the client pitching process — building custom solutions on a prospect-by-prospect basis through the deal process. Nonetheless, a sales and marketing spend representing around 40% of expenditures for a company with just roughly 125 clients as of Q2 2020 is something you don’t see every day.

This consolidated client base was an interesting reveal, too. It would appear that just the 3 top clients of Palantir represent almost 30% of the firm’s 2019 revenue — the top 20 being responsible for 67% of revenues.

This short list of active customers and the disproportionate revenues generated by just a handful of accounts has the effect of making Palantir look increasingly like less of a technology product company and more of a digital services consulting firm. Add this to their extensive investment in the RFP process through custom-built solutions long before a deal is on the table, Palantir begins to resemble a McKinsey or Bain quite quickly, albeit even more secretive. Similarly to the traditional consultants, Palantir promotes its “forward-deployed engineers” on-site at its clients’ offices — integrating client systems with Palantir’s, the company starts to look like a hybrid software/consulting agency — quite a distance from the pure technology company it bills itself as.

So, what’s a Palantir?

With all that number-stuff out the way, let’s have a look at what Palantir actually… does.

Palantir, in a nutshell, specializes in big data analysis and technical systems development. It works primarily with governmental organizations from national health authorities to defense ministries in developing custom software to drive operations via big data insights.

Trying to explain that without sounding boring is a challenge of its own. It gets more exciting in a minute, promise.

Oh, but first a fun fact! The company is full of nerdy references. Offices by the name “The Shire” and “Rivendell” can be found, for example across California. The company name itself, in fact, comes from The Lord of the Rings, in which palantír were “seeing-stones” which allowed their users to communicate with each other or to see faraway parts of the world — a fitting name, then. Don’t even get me started on their counter-terrorism division named Palantir Gotham

The firm was co-founded by a team of 5, 4 of whom were recent PayPal alumni. Of the squad, the most recognizable member remains, of course, renowned tech billionaire and face of the PayPal Mafia, Peter Thiel. The company was largely formed from the talent which once comprised the financial crimes division at PayPal — and this served as one of Palantir’s key entry points into its market. Some of the company’s first clients were members of the United States Intelligence Community (USIC), across financial crime and national security verticals.

Today, the company has 3 main client and service segments, spanning:

  • Financial institutions
  • Law firms
  • Governments

It’s these operations with governmental clients that Palantir is best known for today, primarily for its numerous projects with the US Department of Defense — whose engagements have a large counter-terrorism focus. For these clients, Palantir’s primary value proposition is one of ‘connecting dots’ between agencies. Building powerful cross-platform and cross-agency intelligence databases with massive data mining functionalities has made Palantir the go-to choice for a number of government agencies’ intelligence platform needs.

Palantir offers such a wide range of services that any one specific use case can be hard to discern without digging a bit. So I’ll give an example of the types of use cases Palantir has been used for in the past.

leaked document from 2013 offered some insight into a Palantir-powered solution, here, working with the Pentagon, where Palantir was used to track patterns in roadside bomb deployment and was able to conclude that garage-door openers were being used as remote detonators.

In the finance space, Palantir’s technology helped the Securities Investor Protection Corporation (SIPC) sift through over 20 terabytes of text and financial data spanning 40 years of records, leading to the arrest of a Mr. Bernie Madoff.

Connecting the dots of various agencies, aggregating data from wildly disparate sources, then, appears to be a powerful business! Though — as these recent leaks would indicate — a questionably profitable one.

More broadly, Palantir claims its platform has 4 main uses:

  • Data integration
  • Search and Discovery
  • Knowledge Management
  • Collaboration

Let’s have a quick look at each, one by one.

Data integration

As roughly described above, their data integration capabilities, synthesizing data sources from various formats and locations, connecting it in an ‘object based model’ via the framework Palantir calls “Dynamic Ontology”.

It’s this unique format of relationships between different data structures which fuels the Palantir platform’s ability to uncover ‘relationships’ between otherwise disparate data sources.

Search and Discovery

Through the renowned interface of the platform, non-technical users are able to query all types of data, across satellite imagery, financial records displayed in visual formats.

It’s this complex ability to take that synthesized and aggregated data, and use AI combined with human expertise to visually represent and analyse those connections that constitute’s Palantir’s best known benefits.

Knowledge Management

With such heavy roots in the intellignce community, military and other gonvernment entities, the regulation and oversight required in classifying sensitive and classified information is vital.

In a blockchain-esque breadcrumb trail of each access and modification of any and all data on the platform, Palantir’s software pitches itself as being the most highly-secure entrant on the market.

Collaboration

In connecting expertise across various verticals and analytical focuses, Palantir brings teams together from around the world, offering them the combined data lakes to foster cross-agency, cross-border collaboration.

Individual analysis gets published to a broader network on the platform, allowing others to collaborate and built on top of existing work.

Headwinds

Believe it or not, a private company helping governments expand their military capabilities doesn’t often go without its fair share of public controversy over the years, and Palantir leads the way in that department. From public admission of involvement with the Trump administration by assisting ICE in the deportation of undocumented immigrants, to taking up a Pentagon contract to deploy and monitor military drones (a project dropped by Google following significant backlash), Palantir’s governmental operations have been heavily criticized.

In 2016, it was revealed that Palantir had seen over 20% employee turnover, an eye-catching figure, given its staff of over 2,000 at the time. Later, the US government sued the company on allegations of hiring discrimination against Asian applicants.

Morale and public perception of Palantir were already lacking by this time, following its (unfortunately, quite public) failed attempts in 2014 to create a data-sharing “consortium” across US consumer goods companies, which would have led to an unprecedented bundling of market data across the CPG industry, powered by Palantir insights. Coca-Cola’s subsequent announcement in the following months — among other consumer goods giants — of their refusal to sign onto this inter-firm partnership did little to boost perception of the firm throughout the public sphere, nor industry professionals.

The IPO Path

Talks of a potential Palantir IPO have been long underway in tech and finance circles, largely fueled by intrigue into the inner workings of a company bound to such secrecy.

Investors once believed their chances of ever getting a peek at a fabled Palantir S-1 were little more than a pipe dream, after a 2016 statement by Alex Karp, CEO of Palantir, that the company wasn’t exploring an IPO, namely for the reason that it “would make running a company like [Palantir] very difficult”.

Fast forward to 2020, and investors are finally getting a chance to rejoice in the news of a planned upcoming public offering of the company later this year. Recent sources, though, would indicate that Palantir is looking to follow in the footsteps of Slack and Spotify these past couple years, opting for a direct listing, avoiding the IPO roadshow and fanfare, but also the cash infusion that goes with it.

Through a direct listing, a company doesn’t sell additional shares to the public market, rather just allowing existing shareholders (existing private investors and employees) to sell their stock to public market participants. This decision to forgo the cash-raising opportunity of an IPO is, it too, a surprising bit of news for a company that, well, needs the cash.

Instead, Palantir looks to be opting for the private money approach, having just recently raised over half a billion in new private capital, primarily fueled by Japanese holding company Sompo Holdings. 

In passing up the public money, Palantir is free to raise funds more quickly via private investors, all while skirting the roughly 7% fees collected by the banks underwriting the IPO — all while keeping the possibility to raise extra funds down the road via the public market, but without the bulky IPO process that goes along with it.

Parting Thoughts

All in all, the Palantir public listing is shaping up to be an intriguing one. This first half of the year and, you know… all that came with it… has made 2020 a bit of a quiet time for public offerings, be they IPOs or direct listings (at least we’ve had SPACs!).

I’m excited to dive into the (hopefully) soon-to-come release of the company’s S-1 filing and finally, maybe, just maybe, get a (heavily redacted) glimpse under the hood of the mysterious entity that is Palantir.

Get smarter about business and technology strategy, straight in your inbox.

Join the audience of This Too Shall Pass readers who receive this newsletter each week.

Spotify Audiobooks

The last couple of years have been a busy period for Spotify.

While the company is obviously best known for its music streaming, since 2018, the Sweden-born team has been hard at work setting the groundwork for a true audio empire.

The first entry in the space was of course podcasts around 2016. In suit, Spotify’s last 2 years have been a headfirst dive into the podcasting arena. When they announced they were coming, they didn’t say it quietly. I wrote in Twitter and the Cool Kids Club:

The podcasting space is hotter and more highly contested than ever before. In the last 18 months, Spotify has shown that they are not messing around when it comes to long-form and podcast content.

In early 2019 they announced a deal to purchase podcast network Gimlet Media, and Anchor, a podcast creation platform. This reportedly ran them over $340M. Under a year later, they made waves again by acquiring the huge podcast network “The Ringer” for almost $200M.

A just couple of months after that, the Twitter tech world near-on imploded when news broke of Spotify signing a deal with Joe Rogan, bringing all future and past podcasts exclusively to the platform for a rumored $100M. More recently still, Spotify has announced new deals with Kim Kardashian and Warner Bros, to create exclusive new content for the platform.

All of that to say: the podcast space is a difficult one to compete in right now.

Spotify has been throwing around its weight in the audio streaming world full force and shows no signs of stopping. In fact, while one hand was climbing the podcast production ladder in acquiring studios, networks, and household-name talent, the other was busy getting a better feel of the wider audio landscape, with regards to horizontal expansion.

A few days ago as of writing this, a job posting appeared on Spotify’s recruitment directory for a Head of Audiobooks — leading a team within the original content focused Spotify Studios branch.

Some of the role’s broad focuses are listed as:

  • Developing and executing the book strategy for Spotify 
  • Identifying third-party content to license and package on Spotify 
  • Managing a book Editorial team and strategy
  • Helping develop a long-term strategy for audiobooks

This isn’t the first official mention of audiobooks by Spotify, but it’s by far the most significant. A range of audiobooks already exist on the platform in a mostly unofficial format if you dig deep enough within the spoken word categories. The company has explored some initial entries into licensing longer-form narrated content, but the format has been little integrated into the platform overall.

That may be about to change though.

Streaming Strategy

Content is expensive. Music content, especially the larger an artist is, is exceptionally expensive. It’s for this reason that Spotify, and all major media companies, past a certain size, have begun exploring alternatives to content licensing to fuel new releases on their platforms. The playbook is tried and true: Netflix, Prime Video, HBO, and now Spotify are all getting into the original content game.

Sure enough, at scale, it’s much cheaper to produce and distribute your own content than it is to license 100% of your library. Once initial costs are out the way, no royalties to pay, no license agreements to be renewed, it’s all profit.

Spotify understood this as well as anyone else and turned to new formats to build their catalog of original content, this has primarily taken the form of podcasts. Today, audiobooks look like the next step in the move.

The Audiobook Opportunity

The audiobook market is ripe for adoption. The audiobook market for the US alone has been growing at a roughly 20-25% CAGR for the last few years and is expected to continue in suit — at a size of roughly $2.7B+ as of 2019.

The market, though, is more than 50% controlled solely by Amazon as of 2018. Via their Amazon-own audiobook service and primarily through their standalone Audible platform, Amazon sits leaps and bounds ahead of the next nearest entrants, such as Apple, Scribd, or Google Play.

Challenging the audiobook market, then, is to challenge Amazon. Historically, a pretty bold task.

That said, Spotify has solid — or at least not terrible — odds in the fight.

When it comes to long-form content, particularly spoken word, discovery is a challenge. In a 30-minute listening session, music streaming platforms can get some reasonable initial insights into the type of music you like, and, therefore, what will keep you on the platform longest. By song 10, a not too rough first persona can be gauged from your listening habits, in order to recommend and algorithmically queue other songs and artists likely to keep you listening.

Long-form content is harder. By minute 30, you’re barely done with one podcast, and definitely in the early days of one single book, if not one single chapter. This means that for product-first companies like Amazon and Audible, initial recommendations and cross-discovery engines have been primarily limited to analyzing keywords and category genre. For audio-led companies such as Spotify, their recommendation engines dig much deeper than surface metadata to curate well-targeted recommendations and autoplays.

Already analyzing BPM, tone of lyrics, etc, Spotify already has strong inroads in place to analyze audio content and help craft tailored suggestions. Beyond just analyzing the main themes of an audiobook and recommending others, Spotify’s offering across music, podcasts, and audiobooks perfectly positions it to offer rich cross-format recommendations.

Imagine being 5 chapters deep into Lord of the Rings, and Spotify recommends an associated soundtrack, or suggests a LOTR lore discussion podcast, or chapter by chapter breakdown show. Beyond consuming audiobooks as a relatively siloed experience, Spotify is well-positioned to build more immersive long-term experiences in content discovery. Audiobooks tend to be consumed over days/weeks, this cross-format experience offers an immersive timeframe in which Spotify brings you into an augmented media world, full of spinoff discussion shows, book discussion groups, and soundtrack playlists.

This type of 360-degree consumption experience is of course possible on third party sites, specifically crafted to offer the most immersive reading/listening experience, but creating them would be wildly time-consuming, manually crafted.

Spotify has the perfect technology and positioning to facilitate the process.

Means of Production

Speaking of time-consuming, even making the audiobook itself isn’t quite a walk in the park.

On the surface, the process seems pretty simple: sit down with your book, a mic, and a few days to kill; stitch it all together, export, then hit upload.

In practice, though, the task quickly becomes much harder. ACX (Amazon’s audiobook production network), by virtue of its ubiquity, largely sets the baseline for content quality standards across the audiobook world: dictating specific dB ranges to adhere to, bitrates, credit formats, etc. The process isn’t exceedingly hard to figure out, but most authors will avoid it where possible, typically looking for freelance help on ACX.

Spotify’s audiobook game, though, is one step ahead.

As I linked in the job post at the top, the Head of Audiobooks the company’s looking for works more specifically within Spotify Studios, the original content division, and coincidentally, the one that now controls Gimlet Media and Anchor, a studio/network and a platform both highly capable of developing, producing and pushing out intricate long-form content at scale. No more Findaway Voices needed, Spotify Studios has spent the last 2 years making itself into a lean, mean, content machine.

The infrastructure crossover between podcast production and audiobook production almost entirely matches. While podcast studios of course tick the easy boxes of having the right equipment and quality setup environments, Gimlet Media stands as the perfect entry into audiobooks for its (much more valuable) process knowledge. Having an established chain of production, equipment, experts, industry veterans, and a chunky Rolodex will prove invaluable to Spotify, able to circumvent the growing pains of production at scale almost entirely.

The Amazon Problem

This strategy sounds all good and well so far. Production is no issue, the market is steadily growing, and Spotify has some best-in-class features to differentiate itself in the space. So what’s the catch?

The catch, unfortunately, is a cool $71 billion in cash on hand for Amazon.

Amazon’s leverage power in the book, ebook and audiobook space is, to put it lightly, significant. And that’s no coincidence.

Amazon was born into the book world, building the roots of its empire on a groundwork of books. Amazon has unrivaled negotiating power and gravitational pull with both existing and first-time authors, with the pockets to bankroll any opportunity at a moment’s notice.

Beyond that, the ecosystem they now control today was precisely crafted to offer a self-sustaining process of solidifying their leverage. Amazon dominates the online book market, the ebook market, and audiobook market, proportionally unrivaled. The ability to bundle contract offers to publishers across physical, digital and audio releases of a single piece of content drives down their incremental costs, and for just a little less retained commission, secures them exclusive license rights to almost any and all new book releases in the United States initially, and increasingly further afield.

It’s these slightly pricier exclusive rights (Amazon retains 60% of revenue on exclusives, compared to 75% on non-exclusives) that set the barrier to entry extremely high for any determined new entrant into the audiobook market — and not just new entrants:

For Audible listeners, the yellow band on a book cover reading “only from Audible” facilitates a feeling of access to premium content, but for the rest of the book world, it’s an access barrier. 

It means that the audiobook in question can only be sold through Amazon’s Audible. No other retailers or providers can sell or distribute the digital audiobook, including bookstores and libraries. 

I’m not here to port judgment on the ethics of exclusive distribution rights, rather to show that for the majority of authors lacking an existing large reader audience, capturing that extra 15% on audiobook sales seems like a no-brainer. For most authors, namely the increasing number of self-published writers, where else are you realistically looking at distributing your book/audiobook at any significant volume. Amazon has the audience, Amazon has the infrastructure in place, might as well go full-Amazon — most everyone would have bought from Amazon anyway!

This outsized leverage is the major hurdle to overcome for audiobook entrants, Spotify very much included. Growing a library is hard when your lead competition outlaws the distribution of, well, almost all books. When faced with barriers to entry so high, though, one has to begin to wonder if there’s not another way of doing things?

The Podcast Back Door

While Amazon dominates most markets they enter with their outsized leverage across various verticals, Spotify further dominates in their one specific space: music. It’s this leverage that set the groundwork for their podcast expansion. In essence, attracting users with their more costly music offering all these years, and now attempting to shift that acquired audience towards less costly, owned media verticals, in the form of podcasts. This goes back to the bold list of acquisitions and exclusivity deals of the past couple years, from production studios, to Joe Rogan, to the DC superhero universe, to Kim Kardashian West, to Michelle Obama.

Similarly, this groundwork in the form of podcasts can serve Spotify as their point of most leverage in developing their audiobook offering.

It’s easier than ever before to create books and ebooks at costs lower than ever before. Self-publishing has been on the rise for years.

As such, the majority of ebook and audiobook authors turn to Amazon as the only reasonable destination to sell. While Amazon has long adopted a focus on attracting book-led authors, Spotify is in a position to disrupt the equation by, instead, adopting a focus on attracting passion economy-led creators. In lockstep with the growth of the podcast format on the platform, audiobooks are poised to benefit from the rising tide to float all boats.

The market, today, is undeniably smaller. For authors, adding an audiobook format is a relatively simple process within the entire publishing process. For podcasters, audiobooks represent an entirely new format, but one they’re familiar with from another angle. Where authors look at audiobooks as an extension of their writing format, podcasters look at audiobooks as an extension of the range of products they offer to build a personal (or corporate) brand.

Spotify is already well underway in a mission to attract new podcasters and exclusivize their distribution, in return for more direct monetization opportunities on-platform than they would receive elsewhere. Spotify is in a position to offer a 1-2 combo of affiliate revenue for attracting new paid subscribers to Spotify Premium in order to access audiobooks of creators they already listen to, and the advertising/play revenue already in place for artists on the platform. For podcasters, these audiobooks would less resemble an actual book-style story, but a higher-quality, higher value, story-driven extension of their tried and true podcast format.

Imagine, for example, that the first chapter of a given creator’s ebook was available for listening under the free Spotify plan, just as their podcast already is. For a creator you already listen to and are actively engaged with, you’re rarely not going to at least try that first free chapter. Building off an already warm prospect, the allure of the rest of the audiobook behind only a $10/mo paywall, also granting access to all Premium features already established on the platform, is a strong one. Moreover, it sounds a much more affordable option relative to an Audible subscription, at a 50% higher price point, with extra costs behind that.

For traditional authors, the gravitational pull of the Amazon ecosystem does undeniably remain strong. The revenue gleaned from Spotify on actual plays would be marginal relative to Amazon/Audible — rather, here, using Premium affiliate revenue as a proxy for incremental audiobook releases. As such, an audiobook strategy for Spotify has lower barriers to entry by targeting podcasters and digital creators, compared to those possible in building a fully-fledged Audible competitor, targeting the author and publisher market more broadly. This, while further simplifying the audiobook creation process relative to Amazon, making audiobook publishing more accessible to creators.

Spotify Kids

I mentioned higher up that this recent job posting isn’t the first mention of audiobooks by Spotify. One of the earliest mentions I could find of the format with regards to the streaming company is in the context of Spotify Kids.

In a similar fashion to the popular YouTube Kids app, Spotify Kids is a protected, standalone platform with restricted access and child-oriented content. In the announcement post for Spotify Kids late 2019, they made an initial indication of exploration of the audiobook format in the context of this Kids offering:

As we evolve the Spotify Kids experience over time, we plan to enhance parental control features to allow for even more customization. We’ll also bring our audio expertise to the table with listening experiences that go beyond music—like more stories and audiobooks and eventually podcasts.

The Spotify Kids offering looks to develop their content offering in the opposite way to the standard platform, focusing efforts on story-driven content before moving towards podcasts, but the exploration of the format remains just as interesting.

In 2017, audiobooks for children represented 10% of audiobooks sold in the US, 25% in France, and more than 40% in China. The market opportunity for kid-focused audiobooks is significant. It’s worth noting, in the US, the seemingly low penetration of youth content in the market is likely less an indication of low interest by children, but rather an outsized interest in audiobooks by adults, relative to other markets. In France and China, audiobooks are a much less ubiquitous type of content for adults, largely due to Audible’s aggressive advertising efforts heavily concentrated in the US these last years.

Spotify dipping their toes in the water of audiobook content in the kids space is a strong indication of the ambitions the company has in the audiobook space as a whole. On the kids-end, the opportunity to develop a holistic educational and story-driven content offering is massive, and the existing omnipresence of Spotify in American households would empower its growth in millennial families with young children significantly.

The Ambient Battle

Moreover, the progressive move towards long-form content like podcasting and audiobooks reflects an ongoing shift in the overall Spotify strategy. The production of high-margin original content is a means to an end. Rather than trying to be the lions’ share of audio time, they want to lower the distinction between audio time and time itself — representing a shift towards an ambient media company, filling the silence in the day with something entertaining or informative.

Brett Bivens wrote a great piece specifically about this shift over at Venture Desktop, stating the situation succinctly: 

The real winner in audio over the next 10 years will also need to own an outsize share of something much larger — the consumer’s ambient hours.

To democratize ambient media content, beyond just offering audiobooks, building them into an integral part of your day, a critical factor remains the physical device via which the ambient content is consumed. Spotify’s ambient future wouldn’t have been possible without the rise of the smart speaker and voice assistant services in the last years.

Today, a whopping 25% of households in the US have at least one smart speaker or voice assistant physical device. This is perfect groundwork upon which Spotify can build their ambient media empire. The outsized success of AirPods, also, has heavily contributed to this groundwork also, pushing the “eyes-up” ambient media opportunity years ahead, with over 70% of worldwide wireless headphone unit sales.

If indeed, the real winner in audio will be the one to own the consumer’s ambient hours, Spotify is setting itself up masterfully for victory. Making use of the groundwork laid by audio hardware companies, and developing their offer to expand to long-form, owned media content, Spotify is already underway in expanding their audio empire.

Music was the launching pad for their dominance, podcasts are shaping up to be the second. To take over the fierce battleground of audio streaming services, though, audiobooks will be the next frontier, and Spotify is leading the charge.

Get smarter about business and technology strategy, straight in your inbox.

Join the audience of This Too Shall Pass readers who receive this newsletter each week.

The Rise of the Note-Takers

The first few years of the 2010s birthed a gold rush of development of messaging apps. WhatsApp had been around for a year-ish, Kik launched to immense success in 2010, and Facebook Messenger (then only within the Facebook app) had long been top dog.

Within a couple of years, photography and photo sharing apps were the new developer trend. By 2012, thousands of photography apps had come and gone or were at least on their way out.

By way of the explosion of the mobile games industry, a quick trend in meditation services, and a series of niche utility apps, we fast forward to today.

In the last 18 months, the indie app development world has taken a liking to the note-taking space.

Now in all fairness, this view likely isn’t representative of trends across a more global audience. Within our tech-oriented corner of the digital world, though, the trend is clear.

It has never been more hot to be building a note-taking app. Bonus points if it includes [[bidirectional links]].

What’s most interesting about the space right now, though, is the lack of a clear, widely loved leader. It’s a very fragmented space. Of course, there are dominant players in the form of OneNote and Evernote, but more on them later.

In a sector so full of new entrants and growing competition, why is there still no clear leader? Then the harder question: what will it take for the new notes king to claim the throne?

Note takers, assemble

The rise of the software as a service (SaaS) model over the last 15ish years has been one of the most influential evolutions in the growth of the innovation economy.

At the root of this evolution are the inherent low barriers to entry associated with software development. It costs quite literally nothing to begin developing a digital product (leaving aside, for now, the costs associated with deployment and scaling). Where beginning developing a product has historically entailed high R&D costs, material costs, machinery costs, etc, software is cheap to produce and even cheaper to scale. The marginal cost for each new user of a product, in most cases, is near zero.

Note-taking apps are a perfect example of this.

For developers first building side projects, notes apps serve as a low-cost, simple (all is relative…) technical product to work on. Beyond exploring new technical approaches and interface designs, they’ve also allowed their creators to experiment with new ways to design the input, consumption, and structure of knowledge.

It’s this freedom to explore new formats to store and access your own knowledge that has largely fueled the notes startup trend.

This increased development on the tech side has been amplified by several prominent new influential figures / ‘thought leaders’ (ew) in the productivity and personal knowledge management space. Many of them have been building on top of the rise of new notes apps and developing entire frameworks for their mental models and note-taking process.

Tiago Forte is arguably the most prominent recent figure in this space, with the rapid success of his Building a Second Brain online course, and his range of online content. Tiago’s approach focuses on established tools, but with a rigorous, structured system for his knowledge management.

Meanwhile, Nat Eliason has been a prolific online writer for a while now but has recently seen immense success with his Effortless Output with Roam course, which acts as a deep-reaching intro course to Roam Research, a relatively technical notes platform with some promising and exciting features.

Their efforts, combined with creators such as Conor White-Sullivan — who I recently (jokingly) compared to Creed from The Office — and his growing Roam-loving fanbase, or David Perell and his Write of Passage course have recently brought the art of effective note-taking into the foreground, building communities along with it.

The space is growing quickly, and note-taking is somehow becoming cool. Still, though, there’s no dominant leader that’s widely loved. Evernote leads the pack for now in the personal knowledge management space, but is often criticized (even by myself) for feeling increasingly out-of-date and lacking many of the exciting features of new entrants.

The business of notes apps

The notes app business model is a tricky one to sell.

Most existing companies are based on a freemium model. For personal use, most users can get away with the free functionalities alone, but power users can opt for a ~$5-8/mo package for more technical features and quality of life improvements. Meanwhile, teams and companies have their own enterprise plan, which expands note size capacity, team size, etc. The go-to note app pricing plan roughly mirrors the mainstream SaaS model.

The increasing competition we talked about before, though, does little to help in the way of profitability. Most industries’ profit potential is damaged by the threat of new entrants and the low barriers to entry they face, but the SaaS model only exacerbates this.

We’ve seen the low entry barriers already, and their inherent presence in SaaS products. But it’s their combination with low exit barriers which make notes apps a difficult space to compete in.

Owing to the low long-term investment in technology required to build a strong initial notes MVP, the exit barriers in note apps are almost non-existent.

Looking at this chart from Michael Porter in 1979, the lessons ring true even in a software context. The low barriers to entry in the notes app space combined with exceedingly low barriers to exit signals a market opportunity for relatively muted, but stable returns. Sure enough, new entrants can easily enter the space with little headwinds, safe in the knowledge that should things go wrong, they have very little to lose. Their lack of need for profitability, or even top-line revenue, in many cases, leads to a race to the bottom for non-differentiated or non-niche competitors.

On top of this, several free options are increasingly gaining traction as personal knowledge management services, beyond standard note-taking apps, aimed at both the personal and professional markets. OneNote is a prime example of this, integrating block and line-level linking to allow more networked, rather than siloed thought. OneNote’s bundling with the Office suite and pre-installation on many Windows machines makes it an easy go-to for almost all business use cases.

For competitors unfortunately not packaged with the largest productivity suite in the world, the underlying challenge within growing a note-taking app is almost entirely one of switching cost handling.

Switching costs 

Low switching costs lead to increased churn risk, high switching costs lead to low adoption, and this is a fine line to walk in the notes space. Making the initial jump to a digital notes service as the hub of your knowledge management is a large undertaking.

Notes are an inherently personalized medium. No two people’s notes will look alike, and in some form of subjective esoterism, even highly contextualized notes can’t replicate the entire breadth and depth of their meaning and utility across different readers. Everyone has a distinct mental framework for note-taking, conscious, or not. This framework is largely influenced by the features and limitations of the platform on which they’re taken.

For pen and paper, the limitations are obvious: platform barriers in copying/transcribing, speed, imagery, version control, etc. On the other hand, though, the free-form nature of the analog medium offers full flexibility in style and ease of writing.

For digital platforms, these features and limitations are different, but very much present. The categorization, hierarchy, and depth of the notes I personally take are entirely crafted around Evernote’s features and limitations.

In countering these limitations, many note-taking apps have begun to turn towards Markdown, the lightweight text markup language that’s become near-standard across rich tech editors across the web and software. Markdown offers a homogeneity across platforms and services that hadn’t been found elsewhere. It serves as a basic framework for styling, hierarchizing, and formatting text content. It’s seen increasing usage in new note apps for its ease of global integration, and widespread understanding/usage. This shift, though, towards a more open standard acts as some sort of inverse tragedy of the commons — what’s good for the ecosystem is not good for the company long-term, and yet many do it anyway.

By adhering to a standardized format, many new entrants in the note app market are reducing their ability to fend off the threat of subsequent, even newer entrants. While adhering to a standard format reduces entry barriers for the company, it, in turn, reduces switching costs for the user, lowering the perceived effort required to move to a competing product. When I can copy and paste almost all my notes to a new program (or, more practically, use an online service), I feel very little tied to the platform I’m currently using.

The flexibility and portability Markdown provides is a great benefit to the user, but a serious threat to the platform.

Markdown, though, is a means to an end, not the end itself. It’s easy to integrate as your text editor’s backbone, sure, but Markdown itself offers little in the way of facilitating the smooth transmission of ideas from brain to notes. It’s in reducing as much friction as possible in this transmission that the note app leaders of tomorrow will shine.

Blueprint for a leader

There are a few ways to look at who will become the next personal knowledge management leader. The next king will likely have to tick one or a mix of these boxes:

  • Develop a conceptually new knowledge model
  • Differentiate and specialize their offering
  • Raise perceived switching costs

It’s in this first point that Roam Research has so far excelled. Rather than just serve as a better way of noting down information, it serves as an entirely new (better?) way of thinking about it and exploring it. In Roam’s view, just as brains don’t think in categorized thought, notes shouldn’t either. In using wiki-style bidirectional [[links]], it reduces mental friction in many areas (no more thinking about which notebook things go in, what tags to add…), but currently remains a bit technical for the average consumer (ie. me), right now, adding mental friction in other places instead.

I maintain that the next dominant personal knowledge management leader will heavily exploit AI to further reduce the friction involved in transcribing and contextualizing ideas.

Using Evernote, I add new notes every day from my PC, my MacBook, my phone, and now even a Google Home. Through each of these, I encounter some different form of friction. On the PC/Mac/Phone, I have to consciously format and structure my thoughts in such a way that will be optimal for later discovery in the app. Sure, this method forces me to think through my notes and really break them down, but to be honest, it’s a pain in the ass. It requires me to format my thoughts in a way that’s best for the platform, not for me.

Speaking voice notes doesn’t have all those same issues, but compensates with others instead. It’s almost impossible to format and organize/interlink these notes by voice, and always requires focused time post-hoc to do all the back-end work required to integrate the notes smoothly into the knowledge system.

And that’s the root of the issue, the brain doesn’t think in systems.

Maybe this just sounds lazy or unreasonable, it’s entirely possible. It doesn’t change my outlook, though: the first company to incorporate AI/ML and contextual ‘understanding’ in such a way that actively augments the quality of my knowledge ‘system’, rather than diminishes the flexibility of it, will be an absolutely massive deal.

It will isolate threats from new entrants by massively raising entry barriers to be even remotely competitive. The low development and infrastructure costs I talked about previously quickly go out the window when it gets into AI/ML territory.

This goes hand in hand with the next points of differentiating and specializing the offer, and in turn (significantly) raising perceived switching costs.

There are already companies exploring this space.

Otter is developing a product for the workplace, with AI-powered and voice-focused meeting notes, similar story with reason8, with a focus on natural language processing. Kiroku takes the aforementioned more niche focus, offering a powerful AI-assisted notes service to dentists.

The most disruptive social and media platforms are the ones that bring not only a new community but an entirely new format. The same rings true for the note-taking world. The top disruptors of tomorrow won’t be the ones following the playbook of yesterday — ie. a minimal UI and good tagging systems. Rather, they’ll be the ones reinventing the entire meaning of note-taking.

It’s hard, then, to argue for a single rightful heir to the personal knowledge management product crown. Still, there are clear roads to the throne.

Get smarter about business and technology strategy, straight in your inbox.

Join the audience of This Too Shall Pass readers who receive this newsletter each week.