Mike M. • Call me crazy, but I think the infrastructure/platform areas of tech are exactly where the excitement is going to be over the next couple of years. Many of us close to the industry can perceive a commoditization, but the reality is that for areas like cloud and API management we are still early days. From my own view close to the telco world, I know that we've yet to tap the network side of both markets. In other words, as an industry we've seen phenomenal innovation developed simply on the back of a small device processor or commoditized server. What happens when we have real connectivity? Also, the 'net of things isn't going to flourish without additional investment and innovation from the same providers so many believe are being commoditized.

Chris H. • Ben, I agree commoditization is a constant threat. The API management space blends commodity infrastructure services (i.e. QoS management, authentication/authorization, message transformation) with new infrastructure services (i.e. key management, metering and billing, business activity monitoring, self-service portal) and under-adopted SOA governance practices (i.e. service versioning, service description, impact analysis, change management/notification) to facilitate a comprehensive, business facing use case. The winning technology provider will deliver a comprehensive and interoperable infrastructure platform which integrates into existing SOA platform environments, delivers an extensible service lifecycle management model, offers service monetization and business activity optimization, and flexibly configure services to meet tenant consumer requirements (e.g. data, policy, process). Full-spectrum SOA platform vendors already have the SOA governance, web service security, and service management capabilities built into their products. just a matter of time until the bits are packaged and more vendors jump into the commodity-end of the API management market. Product differentiation will occur based on edge infrastructure use cases, governance process integration, business service monetization, and flexible multi-tenant configuration to satisfy multiple consumer requirements.

Philip C. • I agree with Mike. I am involved in working with arguably the biggest and most advanced IaaS providers in the world, and I will say that they are far from commodities under the hood. Also, as organizations start to transition more and more complex workloads into a cloud model, those differentiators will become more apparent. The fact that the API is the best and most efficient way to manage the services, means that the API vendors such as Apigee will have their work cut out for them as well.

We may well have IaaS commodity services (similar to electric utilities) in the future, but I'd argue that future is at least 5-7 years away. We'll talk commodity when AWS has true peers. Just my $.02

Sam R. • I've already made my views on API Infrastructure clear as you cited in the post, but as a major consumer of Amazon Web Services, DataPipe, and Rackspace let me say that from the point of view of a business that depends on IaaS it's anything but a commodity market.

AWS has low prices and low service. DataPipe has high prices and high service. Rackspace is in between the two. At Apigee we use all three for different parts of our business, and we won't be able to just rip and replace one with the other in the foreseeable future (12 months); nor did we select our provider based on the lowest-cost provider.

While "IaaS" is a homogeneous label for cloud providers, the actual market offerings are anything but homogeneous. Commoditization usually depends on interoperability, which isn't prevalent in IaaS currently. Some companies, like Korea Telecom and Eucalyptus, offer interoperability with AWS, but this is still early. As Philip points out, it will take years before there are enough homogeneous suppliers to call the market commoditized.

In the mean time, don't mistake Amazon's strategy of lowering prices of their services every single quarter as a sign of commoditization - it's a low-end pricing strategy aimed at maximizing market share.

Kin L. • My perception is that the commodization of what we currently view as API management services hasn’t happened. We are still a “little” ways off.

How I perceived the question was not directly are API management services commodotized, but are the vendors in a position to recognize / accept / deal with it when it happens?

With Sam’s original dismissal of Ben’s question, somewhat speaking to this. Sam's view of commoditization may historically be accurate, but the velocity of how it will happen in this new API driven space will much quicker and evolve in ways we have to be agile and open to even notice, let alone take advantage of.

I don't see any problem relying on 3rd party for API management services. I see problems engaging in classic enterprise sales cycle, 1+ year contracts etc. around these services.

To the point in my post, API service providers need to drink same koolaid they are selling. Offer decoupled, modular resources that allow for self-service, self-building of parts / pieces that are commodities or moving in that direction.

A lack of this stance, in an industry, where my competitors are decoupled, scaling as they require, my company is blind, locked-in to services that will prevent us from knowing a portion of what we are doing is commoditized and price has dropped.

To Andy Raskins comment about spotting a commoditized market, and its not what we see here. But everyone building a bunch of services (caching, translation, CDN, security) that are really just shiny new API polishes on things we are doing everyday to support existing web sites. With the velocity we see, it would only take but one player to shift any of what we know as API services to commodities. Let’s not fool ourselves into thinking what all of what we are doing is all that new, it just has a new approach and polish. A lot of the core is pretty proven technologies.

All it would take is one movement in API cool…to say something like WebSockets instead of REST and an open source player to step up….and everything you see as premier API service offering, becomes a lot less valuable.

Or what if Google all of a suddenly opened up Google Console, Explorer and Discovery to any Google Apps account holder so I could launch my API with Google for FREE!

When I get calls from guys like Andy @ Intel asking how to make their API offerings sexier in this space....I guess not all the big boys are in....but something is shak’n.

So my conclusion is still the same:

...That things change quickly, you have to have to be agile, flexible and deliver clear value to stay relevant. This applies to any business, API service providers are not exempt. In today's fast paced, technology driven business climate, I don't think any business leader can dismiss this question, and I think everyone should ask it often...how can I remain relevant?

Great question Ben!

Steven W. • An incisive post as usual Ben!

I’d agree with Sam and Andy that there is definitely very healthy competition and innovation in the sector. It’s genuinely early days and there are a lot more ways for vendors to add value. While price plays a role in terms of accessibility of solutions, the main vendors have significantly different offerings both in terms of architecture and feature sets. We’ve recently added APIs across our whole platform as well as integration with CDNs such as Akamai and CDNetworks which isn’t available anywhere else for example.

Commoditization might also easily be confused with low value, but this isn’t the case. Automobiles are clearly commodities in the strict definition (many vendors, very similar feature sets), however they provide an extremely high value function and innovation continues as diversity of usage evolves. Neither has the price dropped to zero. Our API management systems can handle 10s & 100s of Millions transactions per day per customer, providing you real time control over that data flow – this is something which is needed for APIs large and small, and is very challenging to deliver.

On the larger idea of aspects commoditization coming, I certainly agree with this. Currently the number of Web APis in Programmable Web’s listing is around 4,500 – even assuming there are significant numbers of others which are not listed (there are many private and semi-private APIs) the number is still tiny compared to web as a whole.

While I can’t speak for the other vendors, for 3scale our objective as a company is really to help as many companies as possible unlock the potential of their API – large and small. This is why we already offer our core services out of the box for free and charge only as customers move to higher volumes and enterprise grade services. You could argue this helps "move forward commoditization”, but in that case, I’d say it’s a good thing – it means more APIs. The better tools that there are available to launch and manage APIs, the better since it will move everybody forward - vendor margins may go down over time, but the value both of the infrastructure and the APis provided will go up much more.

Finally, I doubt the depth and quality of additional services vendors can provide on top of the core will top out anytime soon (our "great things to build" list is getting longer no shorter!), so we will be working for a while yet!

Steve
CEO, 3scale networks
http://www.3scale.net/

Alexander A. • Spot-on Steve.

Commoditization: In context of Ben's question and its answers, it's worth it to read this recent post by @msuster :

The Amazing Power of Deflationary Economics http://bit.ly/tnaWDX

Happy Holidays All! -- http://twitter.com/aainslie

Eric K. • Ben, I couldn't agree with you more and I have been saying this for a while - the core technology of API management is difficult to differentiate. I expect to see API management be subsumed by larger PaaS providers - while important it is a feature not a product. Having said that as others have pointed out already there is massive growth ahead in the near term as everybody and anybody gets an API up and running so there's plenty of runway for these (relatively) little guys to make money the next few years before being snapped up by a larger full-line generalist PaaS provider. AND, there's something to be said for building up a practice of API strategy consulting, developer community management, and so, which is harder to replicate than the (relatively) easy to duplicate plumbing technology that makes these things tick, albeit not as scaleable.

Matthew M. • A great post, thanks Ben.

I'm starting to advise all clients that the best way to *build* any application is to start by creating an API that is similar to/compatible with your nearest competitors and then write your web front-end etc. to talk direct to that API. This means that what you are offering your customers is something that you are using yourself every day. It also means that your clients can switch easily from a competitors service to your own (and vice-versa, possibly winning you some brownie points as they leave!) and that you can take advantage of any suggestions that your clients make to improve the API - it's amazing what they think of that you haven't! :)

From a "service user" P.O.V. (we are not, nor do we intend to be a "cloud provider"!) an API is one of the first things I look for in a company if I'm going to do business with them. As a result, I'm all for "commoditisation" (awful word really, but there we go!) of APIs - bring it on!

M.

Eric K. • Matthew that's an interesting observation - copy your competitors style of communication. If words shape thought, then an API shapes business on the programmable web. On the one hand the switching costs ought to be lower when you steal competitor business, but, on the other hand you're letting them define your market.

Sam R. • Matthew's observation is a powerful one, and demonstrates a key principle of competition in technology markets: when there is a clear market share leader, interoperability is a competitive strategy.

We've seen this in network APIs in the past - SMB from Microsoft being aggressively interoperated with by Samba, and the resulting technology bundled in Linux distributions in order improve their penetration of workgroup servers and storage environments where Microsoft held a dominant position.

We're also seeing this in cloud APIs with Eucalyptus building on the success of Amazon's AWS through interoperable clean-room implementations.

In an early market, where volumes are still rapidly increasing and there's no clear share leader, interop may or may not be a great strategy. But if the share has settled out, interop is the obvious way to compete.

Matthew M. • @Eric: What I'm suggesting is that you model the areas of your product which are similar in such a way that it is easy for potential customers to switch over through the copying of a competitors API. You then take this "base" and build on it to give you a competitive advantage by adding the extra features that your customers have asked for, encouraging them to stay with you because of your excellent response to their requirements!

@Sam: If you talk about "The Cloud" to most people, the main organisation they think of (in my experience at least!) is Amazon/AWS. Some may mention RackSpace and fewer still RightScale or Eucalyptus. Based on this, I would suggest (with respect to the other players in the industry) that AWS is the market leader therefore if your systems cannot interop with AWS you may be losing the opportunity to compete.