#9: 15 Years of Market Gaps for Browsers and Browser Extensions
One of the great secular trends of the past fifteen years has been software moving from the desktop to the browser. In 2010, you would install Microsoft Word or Adobe Photoshop and use it entirely offline and on your own. Today, you do the same work in Google Docs or Figma in your browser, always online and collaborative. Personally, the only software on my system other than Firefox and Chrome is VSCode. That’s it. I interface with almost everything through the browser,1 from personal apps like Spotify and Netflix to enterprise tools like Linear and Salesforce. In 2023, the Browser is the universal interface to doing work.
If you presented this knowledge of the future to a startup investor in 2010, the bets to make would have been obvious:
Invest in companies that are porting filesystem/desktop-first software into browser-first software;
Invest in browsers and browser technology businesses.
The first was a great bet. The second wasn’t. In the last fifteen years, basically nothing happened in browser land. But the thesis to bet on it would’ve been so clear:
All major applications move to the browser. Users spend eight, ten, twelve hours a day, exclusively in the browser. The browser becomes the core service and interoperability layer: it doesn’t matter whether you’re on an iPad or a Windows PC, you’ll still be using Chrome to interact with applications, now on the web. But most importantly: the browser is not just a protocol like TCP. These are richly featured, complex pieces of software! There are many browsers with thousands of employees working day and night. No doubt, there will be more browser companies, integrating useful cross-application features, building in everything from website archiving through contextual CRM layers and custom payments gateways…
What’s Wrong with Browsers?
But reality differs. None of this has come through. Even though there are several competing browsers, none of them have meaningfully differentiated on features.2 Worse, for me as a (technical) user, Firefox and Chrome are basically the same today as they were ten years ago; I couldn’t name the last big feature they’ve shipped.3 It feels like there’s an elephant in the room that everyone is ignoring as you contemplate the two big questions in the space:
If users are spending three thousand hours a year inside the browser: can you build a better browser that saves 5% of that time? Surely people will pay to get a couple hundred hours of their life back every year.
If there are hundreds of billions, maybe trillions of dollars of economic activity produced by work in the browser: can you build a better browser that pushes that number up by 1%? If slightly better infrastructure or interoperability can create many billions of dollars of value, surely you can capture some of that.
So what’s up? The only team seriously pursuing this opportunity is The Browser Company building Arc.4 You can explain why there’s little activity in the space: browsers are the most complex modern-day software to build. As a technology effort, it’s too big of a lift for a conventional startup. And while the bull-case is great, the incentives aren’t great overall: needing a better browser is a problem of limited urgency, where marginal improvements don’t yield significant shifts in consumer behavior. Most founders are better served attacking areas where it’s easier to develop a disruptive, 10x better product, like in niche vertical SaaS.
But you don’t have to create a browser from scratch to disrupt the space. For the last fifteen years, we’ve had browser extensions: widgets that you can install that make your browser more useful. There are plenty that you may be familiar with:
uBlock Origin, AdBlock Plus, Ghostery, etc: ad blocking for the browser
Honey: automated plugging-in of coupons and rewards as you shop
NordVPN, Mullvad, etc.: VPNs
MetaMask, Phantom, etc.: in-browser crypto wallets
1Password, Lastpass, etc.: password managers
Loom: in-browser video recording
These are the most popular, and include the most commercially successful browser extensions. As with most things, there exists a long tail of others that are significant in some way, but see only limited use. And when we say limited use, we mean tiny:
These are the Google Chrome Web Store’s top four “Recommended for You” extensions. These top apps have a few million users. By contrast, there are approximately 2.65 billion Chrome users today. On a per capita basis, the usage rounds to zero: basically nobody is using these.
What’s Wrong with Browser Extensions?
Similar to the thesis for browsers, in theory browser extensions should be a vibrant ecosystem of prosperous businesses. The idea of modularly extending your browser to make it more useful bears similarity to installing software on your computer to make it more useful. It’s intuitively easy to get behind.
So, why aren’t there more browser extensions? And why aren’t they more important and better in general? It looks like there are some serious low-hanging fruit, both from business and user experience perspectives.
Browser Extension “Stores” Feel Lousy
Visit the Google Chrome Web Store or the Firefox Extensions hub. Apple’s App Store this is not. Neither of these feel particularly legitimate or well-featured. You can tell from the design, quality of curation, etc. that these offerings aren’t corporate priorities either at Google or Mozilla. These look like afterthoughts. It’s not inviting to the user.
Malware and Privacy Boondoggles
These extension stores are far from the standards of Apple’s App Store. Any browser extension that you install can technically see all your private data, including web traffic and in-browser data. That requires a lot of trust. For the iPhone, Apple solves this by rigorous (1) permission-gating for apps and (2) audits before apps are allowed into the App Store. This is not the case for browser extension stores, pre-publishing review is minimal. And while many extensions are open-source software, there isn’t an audit chain that you can follow to a public GitHub repository. You’re left having to make a leap of faith and trust. And many consumers, myself included, won’t do that for a useful-looking but obscure extension.
Hard to Monetize
Somewhat unbelievably, these extension stores don’t offer a way for developers to charge money for their extensions. The way to monetize is awkward: list on the extension store, then have the user navigate to a separate website, buy a subscription, link up the subscription to the local installation via some credentials… oof! 😵💫 For that reason, most extensions are free. I love free software as much as the next person, but the lack of monetization really hurts this whole ecosystem:
It makes it harder to start businesses whose primary product is a browser extension;
Consequently, most browser extensions are hobbyist efforts that are commensurately unpolished;
It deprives companies like Google and Mozilla of a revenue opportunity;
In turn, that missing revenue means there’s a lack of funding for better moderation and trust, which in turn hurts usage.
Developer Ergonomics are Bad
I occasionally write my own browser extensions to ease certain tasks. It generally sucks. The documentation, community support, open-source libraries to build on, etc. are weak. Toolchains in other areas of software development, e.g. modern React web apps, are much better. Writing extensions feels (unsurprisingly) like a throwback to 2010; there’s a lot of tedium and inconvenience. To me, that’s the main reason why most hobbyist-written extensions feel unpolished.
I’ve mentioned unpolished, hobbyist-written extensions several times above. Bringing this point home: many extensions simply don’t work. They were written years ago and are no longer compatible with updated browsers, or were buggy to begin with. Again, this is an problem of insufficient moderation on these extension stores. Hosting broken software is a great way to get the user to ignore the extension store entirely.
Opportunities for Startups
There are a lot of opportunities to build useful browser extensions. Just by way of example, I would personally love to have any of the following:
Multi-register clipboards so that I may copy-and-paste multiple items inside the browser;
Perpetual as-I-go archiving and summarization of media: make my browser history useful. If I’m reading a Wikipedia article or listening to a lecture on YouTube, catalog it, grab the transcript, and feed it into a searchable index.
Overlay CRMs: let me integrate Salesforce or LinkedIn such that I can automatically see and update biographical or commercial information as I handle my emails or read industry news.
Configure automation of rote in-browser tasks: for example, open up three tabs for Google Ads, Twitter Ads, LinkedIn Ads, set them side-by-side, click on some sequence of buttons within each to get to its campaign summary interface.5 I make these twenty predictable clicks every day. Let me bind it to a hotkey.
Browser-level autocomplete: while GMail has autocomplete, other websites, like LinkedIn or Discord, don’t. But I use the same words and phrases across all of them.
Design inspiration: as an occasional product designer, I collect examples of good design, chiefly typefaces and colorschemes. I would like to be able to harvest those major CSS styles at the click of a button and send to my team, rather than keeping a notepad and pecking through browser tools.
Opportunities for Incumbents
Google and Mozilla, among others, have under-invested the opportunities in browser extensions. I would advocate:
Allow developers to charge customers money in the browser extension stores. This is the key item. Make offering browser extensions as SaaS a viable business model. Everything is downstream of that.
Take a small cut of the fee above. Use this revenue stream to justify a real business division.
Spend the revenues on moderation: make new extensions undergo review, and periodic check-up with new versions. (Perhaps fast-track “verified” publishers so the environment remains friendly to high-speed iteration.)
Overall, create an extension store that feels high-quality and trustworthy. Apple’s App Store is the gold standard.
Improve the developer experience and APIs available when building browser extensions. Make it easy to create business-grade browser extension products and overcome the hobbyist ceiling that is currently so common.
Importantly, these efforts can create meaningful differentiation for these browsers. Apple benefits hugely from having the App Store, not just financially but as an entire ecosystem. Android suffers from the lower-trust, less-polished-offerings on the Google Play Store. If either Chrome or Firefox becomes the browser with the premier extension store and portfolio of superior extensions, that would be a huge win over the competition.
Thanks for reading Loeber on Substack! Subscribe for free to receive new posts and support my work.
You might ask why the browser became the core interoperability layer. I see a handful of confluent reasons, though I’m sure there are more:
The explosion of mobile (smartphone and tablet) devices meant new computing environments where the only common layer was the browser. An iOS app doesn’t run on OS X or Windows, but a web version of the same app can work on all three.
Installing desktop software is generally a tall barrier, and it’s much easier to get consumers to use an interface in the browser.
Technologies like Electron made it easy to export a browser app into a native desktop application when necessary.
There’s a ton of differentiation in the margins, of course. Google Chrome can integrate with your Google account and save your sessions across the cloud. Chrome has user profiles, Firefox has ways to sandbox sessions, Safari blocks ads natively, and so on. But none of these feel like major drivers of use, or significant enough to change stated user preferences.
Not kidding, the last big step forward in browser tech that really blew my mind was when Firefox launched tabs in 2006.
Two important notes on The Browser Company and Arc:
They’re doing a good job!
Arc is built on a fork of Chromium, the open-source core technology that powers Google Chrome, among others. So it’s not a from-scratch build.
There are a lot of single-page-applications that don’t support URL linking to particular interfaces within them, so I have to click the same five buttons to get where I want to be, every time.