Product-to-platform (Part IV)

How to tell if a product has platform potential

In the last three posts, we have discussed three different ways in which firms can turn their products/services into platforms and the various issues that may arise when doing so. To recap briefly, the first method was to host third-party suppliers or developers that want access to the firm’s existing customers (e.g. Apple opened up its AppStore so third-party developers could access its iPhone customers); the second method was to enable existing customers to interact (e.g. Disney+ enables watch parties, i.e. users watching movies together); and the third method was to attract customers’ customers (e.g. OpenTable did this by creating a marketplace between its restaurant customers and their customers).

In this post, we take a step back and address the higher-level question: what types of products/services have platform potential? In other words, we aim to provide some guiding principles for figuring out which products/services are good candidates for becoming platforms.

This is a question we think about a lot when evaluating investment opportunities in startups: we prefer to invest in platforms or products/services that have clear platform potential. For the usual reasons: the prospect of network effects creates higher upside and defensibility. Of course, this exercise should also be quite valuable to founders and executives at companies of all sizes.

The first and most basic requirement for a product/service to be suitable for becoming a platform is that it has a sufficiently large customer base to be attractive to third parties (method #1), or to create enough opportunities for customers to interact (method #2), or to have enough leverage to reach out to customers’ customers (method #3). The platform potential may be there from day one, but it is important for startups to first execute very well on their initial product/service and build up a sizeable customer base before attempting to open the door to 3rd parties, connect customers or attract customers’ customers.

Now we can go a level deeper and ask: what properties give platform potential to a product/service?

It is useful to answer this question for each of the three distinct methods of turning product to platforms that we have laid out in our previous posts.

1) How to know whether a product has platform potential by opening the door to 3rd parties?

A starting point is to think about whether there is the potential for complementary (or add-on) products that add value on top of the firm’s core products. For some products this is obvious. Clearly, there were always going to be apps on the iPhone (whether supplied by Apple or 3rd-parties). For other products, the scope for these complementary products is less clear, and something that needs to be explored. Assuming there is scope for complementary products to add value, the question then becomes whether they should be offered by the same firm or by 3rd-parties.

If a firm’s customers all use its product in the exact same way, then there is probably not a lot of scope for 3rd-parties to add value. However, if there is sufficient heterogeneity in the use cases of the product among various customer segments, then 3rd-parties can be brought in to fill in the gaps left by its product, which is best thought of as a highest common denominator. A recent New York Times magazine on Shopify describes this principle as the 80-20 rule: “Shopify fulfills what’s known in the software development community as the “80 percent use case,” which means it provides 80 percent of the features that merchants need and third-party developers supply the rest — building customized apps for, say, reviews, or discount codes for influencers to give away.

It is quite obvious that this principle applies to the most successful examples we have mentioned in the post on opening the door to 3rd parties: the iPhone, Amazon Web Services, Shopify, Twilio, Wix, Zoom. A smartphone is an exceptionally malleable product that can be used for a million different things. That may not have been immediately apparent to Apple, which is probably why the iPhone was launched in 2007 as a 100% Apple product. Within a year, Apple realized it could never hope to think of and produce all the apps that its users would be interested in, so the company opened up APIs and the App Store. Similarly, given the large number and variety of companies that use AWS and Shopify, it is pretty clear that there are numerous opportunities for 3rd-parties to complement Amazon’s and Shopify’s baseline offerings.

This is arguably the most powerful reason why opening the door to 3rd-parties makes sense. Any firm selling software that provides a core service to a large number of customers should consider opening up APIs for third-party apps/services that may want to provide additional features on top of its core services.

But sometimes the potential for opening the door to 3rd-parties can be driven by more opportunistic considerations. If a firm’s product/service’s customer base is particularly large, attractive and hard to reach otherwise, it can sometimes make sense to open the door to 3rd-parties to advertise and/or sell products or services that are completely unrelated, simply because the initial product/service offers a unique and convenient channel for accessing those customers (and this channel can be monetized). Examples:

  • Many airlines engage in the annoying practice of having flight attendants advertise credit card and hotel partners over the in-cabin speakers, especially when passengers are trying to sleep. Some even place ads on the back of each seat’s tray table. Similarly, Uber and Lyft have recently begun allowing advertising inside their drivers’ cars.

  • Hair salons can offer a captive audience to advertisers and 3rd party suppliers of all kinds of services, which may have nothing to do with hair styling. For instance, some of our students proposed that hair salons should allow companies come in and administer short surveys to their clientele, who are often stuck passively sitting in a chair for half an hour or more.

  • Tesla and other autonomous vehicle brands in the future:  offer the ability for entertainment and information providers to sell to users via the screens in these vehicles.

2) How to know whether a product has the potential for connecting customers?

This is equivalent to asking why would a product/service’s customers want to connect with one another? Based on the examples discussed in our post on connecting customers, there are several possibilities:

  • Exchange tips and best practices related to your product. This is a natural fit and can add significant value for more complex or fast-changing products/services (e.g. Wolfram Mathematica, Tesla’s electric cars). For other products/services, such customer forums may add little value (e.g. BMW’s exclusive community for drivers, Samsung’s online community for discussing various things related to Samsung products).

  • Provide services that are adjacent to your product. We have discussed the example of Intuit QuickBooks and its “find an accountant” service. Extrapolating, this could work for any product where some customers can provide services to other customers based on the product they are all buying. For instance, fitness wearables (e.g. Apple Watch, FitBit) are used by both consumers and personal trainers and could be extended to matchmake between the two. The wearable provider could then allow users to share their fitness/health data with the personal trainers they have chosen, thereby enhancing the interactions between those two groups. Of course, this opens up the possibility to charge trainers for a “pro” subscription—software tools that would let them access clients’ data to oversee activities and progress. Apple seems to be moving in this direction with the Apple Fitness+ service for the Apple Watch, announced for late 2020.

  • Shopify could help with user discovery and get more direct network effects across its massive merchant base without creating any concern merchants are being commoditized. To do this, merchants could opt in to allowing other merchants to advertise after checkout to their customers, and Shopify would build the recommendation engine which would help optimize the set of merchants that would be shown among those willing to advertise. The more merchants Shopify has, the more cross-selling advertising options there are for merchants (a direct network effect across merchants). Moreover, the more people use this feature to click on other merchants' products, the more data Shopify would collect to optimize its recommendation engine (a data network effect).

  • Engage in social interactions around your product. In our earlier post, we discussed the example of Webkinz digital toy avatars. Extrapolating, this can work for any product/service whose use case has a social component, so that customers would enjoy using their respective products together. Games and toys fit naturally. But it can also work for services like online dance lessons (e.g. Steezy) or fitness classes (e.g. Mirror, Peloton), which can be extended to enable competitions/challenges (online or offline) among users of the service.

  • Of course, a special case of social interactions is dating. Which leads to the interesting question: what products/services are good candidates to be expanded into dating services? We’re not too sure about smart fridges (recall Refrigerdating) and air travel (recall Virgin’s inflight flirting system) but it seems to us that services like Netflix, Kindle, Spotify, Peloton and Steezy would fit the bill because movies, books, music, fitness and dance should be good indicators of dating preferences.

  • Engage in secondary market trading. This can be a particularly valuable way of connecting customers for products/services that offer access to non-traditional assets that are not very liquid. For instance, AngelList allows its members (who are accredited investors) to invest in startups via investment syndicates created by lead investors. Given how illiquid startup shares are and given how many companies AngelList syndicates deals in, AngelList could expand its service to enable trading of startup shares among its members.

3) How to know whether a product has potential for reaching out to customers’ customers?

As we discussed in the post on reaching out to customers’ customers, this is mainly relevant for B2B products. Going further, this potential exists when there is significant scope for a firm to help its customers better interact with their existing customers and/or reach new customers. So this works well when the firm’s customers are small businesses without deep expertise and established customer bases, e.g. restaurants for Open Table, small online merchants for Shopify, or potentially consumers of any brand that wants to help them resell its product. Furthermore, the current product that the firm offers its customers must be the basis of the products/services they offer to their customers and which would benefit from being aggregated in the same place. For example, Shopify’s software tools are the basis of its customers’ websites, where they allow their customers to browse, order and pay for purchases. Clearly this creates the opportunity for Shopify to offer a common payment system and a way to track deliveries.

The Shopify approach of going halfway on the road to reaching out to customers’ customers can be adopted by B2B companies offering similar services in different verticals.

However, we don’t think it necessarily works for any B2B software provider. For instance, Airhouse provides fulfillment and logistics services to direct-to-consumer (DTC) companies. One could envision Airhouse creating a website where consumers can browse and order from DTC brands that use Airhouse’s fulfillment and logistics services. However, we view it as a much less compelling opportunity than for Shopify because fulfillment and logistics are a less natural lynchpin for creating a one-stop shop for consumers than software powering consumer-facing websites. Similarly, Prodigy provides software tools to car dealerships that allow them to streamline the sales process for their customers (car buyers). Prodigy could create an app for consumers to find the nearest dealerships that use its systems and begin browsing available cars and sales terms and conditions, but that seems of limited value given how infrequently consumers go to a dealership to buy a car (contrast with ordering food or buying things online).

To conclude, the approach we recommend for companies and investors is to go through all three product-to-platform methods, generate the best possible incarnations of each method for the product/service under consideration, and evaluate how compelling each of them is.


If you enjoyed reading this, sign up now so you will receive our posts directly in your inbox as they are published.

In the meantime, tell your friends!