I’ve recently discussed the problems with nearsightedness when designing APIs through the comparison of a OAS (Swagger) API to a hypermedia API. These discussions have been very technical, targeted at an audience of API designers largely without addressing business and economic ramifications. In this post I’d like to take a step back to remove my technical hat, and talk about the business, economic, and human benefits of supporting and maturing the proliferation of hypermedia APIs. I’ll go into some differences between these two options from a business perspective to demonstrate the massive value of hypermedia APIs. I will end with a little on a related topic, the undue influence on the direction of technology from the venture capital backed world of hyper growth.
Your API is your product.
Let’s talk about the elephant in the room; I would like to address the extremely common misconception that an API is nothing more than your UI app’s gateway to data. The term REST as the industry understands and uses it, reduces the value proposition of your API development to little more than a gateway between your product, and how you store your product’s data. It may provide some functionality, enhance performance, and shape the data in a way which is beneficial to the UI app development team, but it provides no value in itself. Not only does this go directly against the path of progress towards the API economy, but it wastes the opportunity to save time and money on redundant parallel effort. You have driven initial costs up with duplicated development effort, and maintenance costs also go up due to similar bugs in many places. Perhaps the most critical effect of this mistake is you have almost certainly increased the time to market for your entire solution.
Your customer is everyone outside your API.
It may be difficult to look at your API as anything more than a means to provide your real products with the data they need to create value, but this thinking is guaranteed to hit your bottom line in a big way. The API is your product, and anyone interacting with it is your customer. The internal team developing your new mobile app? Your customer. The group responsible to maintain your web application? Your customer. The outside parties looking to utilize your service, without having to use your mobile and web apps? Your customer. Each of these groups share the exact same goal, they want to utilize the functionality your API provides as quickly and easily as possible. If your own employees, and your customers all share the same goal, you are missing a tremendous opportunity to capture efficiency gains by making your API easier to use. Anyone using your API regardless of their affiliation with your company wants to learn as little as possible about your API to meet their goals.
Solutions like OAS offer short term benefits to your product. Developers can quickly get up and running, there is ample documentation explaining the ins and outs of your system, and users can utilize this to leverage your products very quickly. The catch is these solutions offer extremely poor performance in the long term. Over time your customers will need to constantly maintain their code, continue to need to read and understand your business models to meet their goals, and their quick solutions turn into a nightmare of legacy code to fix. The ample documentation you thought was such a victory becomes a high barrier to entry as your service matures. The result is an extremely dissatisfied customer, one who won’t refer your product to a friend or colleague, who is only looking for a better opportunity to present itself before they leave your product in the past. When they are gone, they aren’t likely to come back, they already know how bad it is to use your product.
When you decide to use a product like OAS to form the foundation of your API, you prioritize your needs over the needs of your customers. The short term benefits of OAS disappear quickly, but the long term negative effects to your business and brand will be extremely costly to remedy. A product which puts the priorities and goals of the customer front and center will drive referential sales, creating buzz and goodwill in the marketplace surrounding your brand. If you want to create long term goodwill and revenue security, you need to prioritize how people feel about using your product.
You can’t sell a mega product.
Technology is forcing businesses to change how they sell their products to the market. The concept of selling an entire package of solutions is quickly yielding to selling smaller incremental sets of solutions which can be independently acquired and used when the customer needs them. Digital products are quickly becoming commoditized to the logic and value they add to your customers’ business processes. If your business model is not changing, you will likely soon find your target market has dried up. Inferior, but modularized products will take the place of your products as customers learn to carve out the functionality they need without unnecessary costs or complexity of larger bundled solutions. Every organization is in a race to the bottom on cost to provide products and services, if you force a customer to buy products they don’t need, you are continuously inviting them to seek alternative products. When you promote goodwill and engagement with your customers, the easiest sales channel of current customers gets even easier to sell enhanced services. Your customers are more likely to buy additional products from you when they are already satisfied with their current solutions.
You can’t sell a mega product, but you might be able to sell a customer all the parts of one. Fixed API designs like OAS make segmenting your products difficult and unintuitive, while requiring a lot of management overhead which cuts deeply into the margin for the product. The dynamic design of hypermedia APIs allows your product to be segmented naturally. This enables marketing initiatives to directly target specific functionality and customer pain points, while requiring very little additional overhead to reduce the profits. If your segmentation isn’t intuitive and it isn’t easy to determine where the functional and license boundaries lie, your customer experience suffers dragging your future sales potential down with it.
Your business probably isn’t hyper growth.
It is difficult to look at the success of companies like Facebook, Netflix, Amazon, and Uber and resist the temptation to copy the way they operate, however its very likely the needs of this niche market do not match your organization or your industries’ needs. The move fast and break things mentality of Silicon Valley and other venture capital funded startup hubs pairs extremely well to the short term benefits of API designs like OAS. In the venture world a problem two weeks away can feel like two or three lifetimes. Companies who intend to hyper scale for acquisition aren’t concerned with their customer’s success in 2 and a half years, because after they sell in 2 years it will be someone else’s problem. Google and Amazon build, try, and sunset so many products that it’s folly to waste time concerning themselves with long term benefits to themselves let alone customers. If you are reading this then there is a very good chance your needs don’t align to such short term goals. Trying to operate using the same tools and methods as these hyper scale companies will do a disservice to your customers and your brand. Your business model likely is concerned with your customer satisfaction in two, five, even 10 years. Hyper scale companies have, and continue to develop, good tools for their needs, my advice is to look carefully to see if those tools fit your needs, because it’s likely they don’t.
Hypermedia APIs are simultaneously an extremely proven design, and an unexplored frontier. The internet itself runs upon the very same principles of a well-designed hypermedia API. Developer tools for this space do currently lag behind the alternatives like OAS. Fortunately, they have the same potential for speed to market, prototyping, and integration tools. Investments in developing hypermedia APIs and the tooling around them are investments for the future on the scale of decades. Hyper scale companies have created tooling which prioritizes their goals of short term gains, if your company is not primarily interested in short term gains then it is up to you to create the tools which prioritize the long term benefits to match your goals. The long term benefits to your company and your customers of developing a hypermedia API have no equal, there isn’t even a good alternative to compare. If your business is concerned about long term market sustainability, revenue, and customer retention you should be looking into hypermedia APIs.