I want to talk about how the dynamics of crypto infrastructure have changed the way we think about a feature versus a product. There are two points I want to make here:

The first is that the composability of Web3 allows features to find adoption in a way that we typically only see with products, obscuring investor/founder’s ability to distinguish between crypto products and crypto features.
The second is that tokenomics can further obscure the typical filters used to differentiate a feature from a product.

First, let’s look at some of the questions that help define what is a product versus a feature?

Is this tool a standalone product that customers would purchase?
How valuable is this functionality as a standalone application?
How much of a nice-to-have vs a must-have is this functionality?

Essentially, it boils down to how valuable is the problem you are solving, and whether or not this problem is solved in a way that allows you to continue growing and expanding the product you have created to solve adjacent problems. A feature is a characteristic or function of a product that adds value, but only incrementally. The size of the problem being solved is also a reflection of the market size, which is a reflection of how much of a fit it is for venture capital.

An example can be in the personal wealth management space in Web2 Fintech. Is a dashboard to track all your bank accounts a feature? Probably. What about a dashboard that allows you to track all your subscriptions? Probably. What about a dashboard that allows you to do all the above, allow immigrants to remit money back home, and get their first mortgage? If nothing close to that exists out there, that probably starts to look more like a product with a thicker value proposition.

Let’s talk about 3 famous examples of the Feature v Product debate, as described by Yannick Oswald:

Dropbox
Calendly
Spotify

The famous example that ignited the debate was when Steve Jobs called Dropbox a feature back in 2009. To the Dropbox founders, Dropbox is a product, but to Steve Jobs, dropbox is a feature like iCloud. What Steve didn't think of was what was behind Dropbox's atomic unit of backing up files. Dropbox created a platform that makes all of your digital content (images, videos, documents, etc.) easily accessible across different situations spanning different people (sharing, collaborating, etc.) to various places and technologies (Mac, Android mobile, etc.). To

At first, Calendly feels like a feature on top of our digital calendars. But there are so many people out there who find use in such a feature that it was a great wedge to build a larger independent company. Currently, some 10 million people are using Calendly every month. And, with the pandemic, the number of meetings we now need to set up has gone up. All of the serendipitous and impromptu encounters we used to have around an office, or a neighborhood coffee shop?

In its early days, Spotify launched a killer feature that lets users share songs and playlists without worrying about who is and isn't currently a member. The atomic unit of Spotify was and is allowing users listen to any song they want without having to buy it, of course. So, playlists were only a social feature, but a brilliant one that made the platform attractive against competitors such as Apple. Features can make products great and get them ahead of rivals. It is important to note however that Spotify was more than just a playlist of songs. There were many startups that tried to go to market as a ‘standalone’ playlist creator, that failed, because it did not own distribution and streaming.

How does this apply to Web3?

Web3’s never-before-seen hyper composability has helped obscure the distinction between a product and a feature. In Web2, a new product with only an incremental feature, would face resistance in adoption, given the switching costs needed (i.e. you would not switch banks just to use a personal finance dashboard). In Web3, the interoperability of products meant that new features could be built on top of existing dapps, ensuring ease of adoption within a single release, but failing to address the core question of whether something was a product or a feature. Users are able to shift between products very easily and get the exact same experience and service as on the previous product with an additional feature benefit.

This allowed many Web3 companies that were really building incremental improvements and features to claim a high level of user growth, despite fundamentally having a narrow product focus that was difficult to expand from.

An example of this can include auto-compounding vaults on top of Uniswap. Auto-compounding vaults are incredibly useful for users looking to increase yield without having to manually compound yields, at the expense of repeated gas transactions. However, this is arguably a useful feature rather than a standalone product.

Another example of this can include semi-custodial exchanges. As MPC technology matured at the same time in the aftermath of FTX, ideas around building a semi-custodial exchange that allow users to interact with a centralized exchange through a non-custodial wallet began to proliferate. Such platforms, utilizing a semi-custodial interface that is likely using third-party technology, is a feature rather than a product, especially given the likelihood that this tech and process will become a market standard.

I had raised earlier that there are many great products that can start out as features. How does this apply to Web3? It can be good to ignore the question of products vs features in the context of a fast-growing market. It is possible to employ a first-mover advantage with a feature to occupy mindshare and distribution in a nascent market, and accumulate the resources and customers to continue building a sticky product. This is to say - there are instances where investing in features in a fast growing large market can be beneficial. Going back to semi-custodial exchanges, they are again an example of this trend that may benefit from the boom as first-movers of semi-custodial solutions gain some headwind. It is then entirely possible that a semi-custodial exchange born today, through good/fast execution can become a new overall market leader.

There are a few great stories to understand how this range has played out.

Convex

The entire value proposition of Convex is that it enables bribery as a service entirely around a single platform, Curve. Essentially a voting mechanism built on top of Curve, it was able to ride on top of the importance of Curve and have its own Convex tokens be an important source of governance for liquidity. Efforts by others to replicate Bribery as a service for other platforms have largely stagnated.

Blur

The entire value proposition of Blur is that it gets to lean on the volume and liquidity on top of marketplaces like Magic Eden and Opensea. Blur was essentially a trading UX that was able to build its own liquidity and generate a liquidity moat against the marketplaces it was drawing liquidity from.

Aave

We can see how limitations to composability can affect how to think about the feature vs product debate across different types of crypto projects.
Across Aave V2 and V3, liquidity is fragmented, with one potentially arguing that each version of Aave should serve as a standalone feature, while Aave as a collection of pools LPs can interact with, serves as the product

As Yannick mentioned, “The product needs to be so good that they can acquire a big piece of the market for their atomic unit and rapidly extend their family of features and products over the coming years. It is this dominance over the atomic unit which enables these companies to build category-defining businesses in the long term. As a result, incumbents and newcomers should quickly face difficult-to-cross moats in their attempts to compete.”

It can also work the other way. Ideas that were products can also become features as the problem initially tackled shrinks in urgency. A potential example of this in Web2 is bill payments. Where previously bill payments were a problem to solve requiring a standalone product, in this day and age where there are so many tech-enabled and plug-and-play solutions, bill payments became a feature rather than a product.

An example of this happening in Web3 can include the many platforms that have scaled horizontally quickly to become the de-facto platform offering a range of services that ultimately allow them to be sticky, revenue-generating core infrastructure like Biconomy.

Why feature-type platforms may have more longevity in Web3 than Web2

In Web2, tools that were more features than products struggled to monetize. There was a feedback loop, typically around Series A, that existed where it became clear that product expansion was difficult, or monetization became difficult as the feature was not sufficiently interesting enough to get a larger segment of customers to switch over to a new product with the feature entirely.

In Web3, one key economic innovation was the ability for tokens to be used as a reward mechanism to sustain open-source protocols that were previously incredibly difficult to economically incentivize development. This essentially meant that as long as you could prove utilization or simply hyped it up, there was an ability to monetize your activity through the sale of tokens. This led to the joke that crypto featured hype as a business model, as tokenomics became an obscuring force that required us as an industry to adopt new measures of what it means to be a product vs. a feature. Obviously, there are plenty of teams that are building and solving real problems in crypto. There are plenty of crypto businesses with real product market fit, with significant 9-figure fiat revenues. However, given the ease of monetizing tokens, many companies simply do not need to face the same issue of understanding if they are a product versus a feature. Furthermore, given the speed at which the industry grew during the boom years, it hardly seemed to be an issue, as many tokens launched managed to hit mouthwatering valuations even at early stages of product development.

As such, we should expect and continue to expect many more well-capitalized feature-type platforms exist in Web3 than Web2.

How does this affect the way we think?

As Founders:
- Moats, particularly around switching costs, that once protected Web2 products are no longer viable in Web3. Due to composability, teams operating in Web3 are going to have to find new ways to develop and maintain their user base.
- Product ecosystem thinking may become more prevalent as businesses initially building features grow out a range of features that allow them to stand as products.
As Investors:
- Thinking through whether something is a feature versus a product has become a lot harder in crypto, with the key being able to understand what features are more likely than others to retain and occupy important mindshare. With certain tailwinds, feature businesses may leverage fast growth to be something much more.
- Tokens have also made it far easier to make money on something that simply has a lot of marketing.
- Long-term oriented investors are more focused on hype-resistant business models that ultimately lead to products versus staying as a feature.

Thanks Daniel Fallon-Cyr, Karthik Rajeswaran, Sunny Singh, Brandon Galang, Mihir Kulkani, Carl Hua, Sehaj, Will Reid, etc for comments/input. Input does not imply endorsement.

Messy Problems

Product vs Feature: The problem in Crypto