Broad and Narrow Internet

But this selection, this narrowing, is once again not a diminution, but an increase. By restricting or selecting, something new that was not there before comes into being. The process is like sculpture, in which a thing comes into being through something else being pared away.

- Iain McGilchrist, The Master and His Emissary

This essay is the spiritual offspring to Chris Dixon’s Two eras of the internet: pull and push. Dixon begins his essay, “The evolution of the internet is an extremely complex topic. Sometimes it is helpful to find broad patterns that make it easier to understand.” This essay attempts to do the same.

Two types of products exist on the internet: narrow and broad.

Narrow products are focused, and don’t care much about offering something for everyone. Broad products seek to service most, if not all, internet users. For instance, Amazon seeks to be the ‘everything’ store, a very broad mission. Narrow products seek specialization. Chewy wants to be the best online pet store, a narrower mission. Broad products are sometimes called ‘platforms’ and often rely heavily on network effects. They are also often marketplaces— they match sellers/creators of content, rides, knowledge, and other products or services with consumers. Narrow products are less reliant on network effects. Narrow products deliberately contain fewer features.

These terms, broad and narrow, are the poles of a continuum. Amazon is broader than Chewy. Etsy is broader than Artsy, but narrower than Amazon. StubHub is narrower than eBay. Brilliant.org and Duolingo are narrower than Coursera. Narrow and broad also refer to several properties of internet products. The terms can just as usefully describe the array of features, or the scale of information managed by the product, or the scope of the mission.

During the 2010s, the maturity phase of Web 2.0, the dominant trend was toward broad products— broad feature sets, broad content variety, broad mission, broad markets. The Web 2.0 companies that succeeded at this time are almost all broad. It was an era of winner take all. The definitive examples are: Google, Facebook, Instagram, Twitter, YouTube, Amazon, Twitch, Uber, Airbnb, Venmo, eBay, PayPal, etc..

Another way of thinking about broad products is imagining the quantity of information available during a typical interaction. A single Google search returns millions of results, an Amazon search returns thousands of products, a YouTube search returns thousands of videos, a Twitter or Instagram feed is essentially infinite, an eBay search returns thousands of products.

Most broad products solve this quantity of information challenge with sophisticated algorithms, which filter and sort the information for the user. Despite their filtering and curation, broad products still give the user lots of information options. Narrow products have an opportunity to be even more restrictive and focused. A few examples:

• Finding the answer to a specific question on a broad search engine like Google is inefficient. No one needs a million results, or frankly, even an entire page of results. A much narrower product would be more efficient for most queries. Asking the same question to an AI LLM (Large Language Model), and receiving a natural language reply is a narrower solution.
• On broad social networks, the algorithms are trained not only to filter and sort content, but also to capture the user’s attention. Narrow social networks have fewer incentives for undesirable user attention capture. For example, BeReal narrows several features of a ‘regular’ social network. It only allows posting at a specific time, it limits the type of post to a single format, and it limits the social graph to people you actually know. It's narrowness gives it space to compete with the broad incumbents, and limits the perverse incentives.
• On broad marketplaces like Amazon, the quantity of nearly indistinguishable products is vast. A search for 'dog food' returns over 4,000 results. Narrower marketplaces specialize their service to users and undertake more of the decision making on behalf of the user. Narrower marketplaces curate their goods more discriminately, easing purchase decisions simply by reducing the number of arbitrary choices.
• In many ways, Notion is a more broadly capable application than Google docs. But it's document creation feature is narrower. For example, there is a finite set of text formats. Users cannot pick their own font, or change the body font size. The result is that editing text in Notion is simpler than it is in Google docs, and it looks better.

I am not the first to describe some of these ideas. Sometimes the concept I am describing, a shift from broad to narrow, is called ‘unbundling’, or ‘verticalizing’. Over the course of history, it seems that this pattern– from broad to narrow– recurs infinitely. There is a famous quote from Jim Barksdale that describes this phenomenon, "... there’s only two ways I know of to make money: bundling and unbundling.".

Jeff Jordan and D'Arcy Coolican support this pattern in their essay Platforms vs Verticals and the Next Great Unbundling, “The moral of the story is this: In all but a few circumstances, the broad horizontal verticals eventually break.” Broad internet products are susceptible to more focused, narrower competitors.

The internet will continue to host both types of products. The tension between them will be interesting to watch. Web 2.0 seems to have been won by broad products, but if history is any guide, narrow products are due for a comeback.