Because of some conversations I had and a talk I gave regarding digital business modeling last week, people reached out to me with questions about the differences between Platforms as a Service (PaaS) and Software as a Service (SaaS). In more detail, the questions were about my views regarding the strategic implications for value creation, business model development, pricing strategies, and when to start thinking about these implications. Because the question came up almost a dozen times, and I can only process so much coffee, I decided to write it down in this mini-blog about a phenomenon which is called XaaS, Anything as a Service.
The Strategic Importance of Choosing the right XaaS Model
As those who know me can tell you, I will passionately argue that this is not a simple semantic discussion. Making the right decision regarding your product or service is of tremendous importance. It is not a tag in a business model and certainly not a buzzword. The development of a software product, business modeling and pricing, and branding and marketing is a holy trinity. These components go hand-in-hand, not in sequence.
First, the latter part of the abbreviation, “aaS,” relates to a service-oriented structure as you can see in the product image below. Microsoft Office is a great example of this. It makes complete sense in light of a pre-WWW world but because of how The Web broke distribution channels (if you are unaware of this, please read Ben Thompson’s piece on Aggregation Theory) it became much easier to go directly to the consumer as a producer. How do you bring a superior user experience and stay in touch? Through an online service! You can have an ongoing interaction with your customer, and thanks to the internet (more specific, the web, they are not the same) you can directly push new value (e.g., a software update) to them.
Garbage-collection as a Service
The first part of the abbreviation is about what it is that you are turning into a service. As I always like to point out, this is not new. Every Wednesday, my local government provides me with garbage-collection as a service. Or the local gym provides me with a gym as a service. I pay monthly fees for a subscription to a service.
In the world of bits and bytes, there are three famous one’s Infrastructure, Software, and Platforms. All as a service. As you can see in the slide above, we make a distinction in SaaS for Office 365 and PaaS for Google’s Gsuite. Although I don’t go into depth in this blog about Infrastructure as a Service, it holds historical importance because it kicked off the XaaS hype through Amazon’s Webservices. If you want to know more about this, please watch this amazing fireside chat with Michael Skok and Andy Jassy about the history of Amazon Web Services.
Last but certainly not least, within digital products, the service model works specifically well because of the effects of scaling software based businesses as described by Peter Thiel in his book zero-to-one. Thiel argues that one of the powerful things that software brings — especially online — is that going from zero customers to a single customer is hard; the software is difficult to build and costs a lot of time and money. But if you do it well, scalability costs you almost nothing. Scaling is so cheap that it is almost hard to determine how many digits after the decimal point it costs you to add another user. Take Twitter, for example. If one would rebuild the platform, it would cost a tremendous amount to serve a single, first user, but almost nothing to serve the second, third, hundredth, thousandth, and millionth. If a percentage of these users become paying customers on your platform for a recurring fee, your income rises exponentially while your operational investment stays almost linear, and with well-designed software, it might even go down over time. This also explained why it is easy to keep the the service that creates the actual value free and look for more profitable form of revenue like Google Search does through AdWords.
SaaS versus PaaS
With SaaS (Software as a Service), you are getting software. Take for instance, older versions of Microsoft Office. The thing here lies in the word “getting”. You are receiving it through an -often online- channel, but it runs on your machine (or in the case of business software on your servers, like Microsoft Exchange). This is a smart model to choose if the value of your product lies in people having access to the software. In a corporate context, this makes sense. You can keep your Excel files locally and your email, too. Microsoft’s software is known to have many bells and whistles. This is in line with the demanding nature of corporates wanting a lot from a single product, and why Excel is still a success.
With a PaaS (Platform as a Service), the user is going to you. You are not “getting” the software. Google’s Gsuite is a great example of this. You go totheir website (docs.google.com) and you consume their service. The great thing is that Google is in full control, they can push whatever they want to you, offer you unlimited storage and an amazing user experience. Did they make a mistake in the software? They change it, maybe even the same day. This is arguably the reason why Google Docs doesn’t have many bells and whistles. Do you want a bell or whistle? Use their platform APIs!
A SaaS or PaaS solves different problems for different end users. It is extremely important, though, that you choose upfront which strategy you will use. As demonstrated on the slide, it is extremely difficult and costly to change course. It will also impact how other products and services are formed within your company. Take for example Google backed moon shoot Project Loon.
Project Loon — Balloons to Provide Access to The Web
Project Loon is a Google backed project which is an example of a logical, physical business spinoff which goes hand in hand with Google’s ecosystem which is based on a PaaS business model. On the Project Loon website, you can read that the balloons are “designed to bring Internet connectivity to rural and remote communities worldwide.” An example would be Africa. In central Africa, only 0–9% of people have access to the Internet (as of 2015). But access to computers and smartphones per person is higher. There are not necessarily more computers, but the numbers of people who can access one (through internet-stores, schools, or libraries) are significantly higher. If Google can increase access to the Internet, it is tapping into a vast number of potential future customers: a simple Google account with access to Google Docs costs the consumer exactly nothing. The PaaS model suits their competition against Microsoft Office.
Internet users in 2015 as a percentage of a country’s population. Africa clearly shows as the largest single area behind the digital divide. Source: International Telecommunications Union
New and Old Models
Because of the widespread use of service models, we often forget that there are also very effective older models for software monetization, like Oracle’s (none-cloud) licensing model and more future-focused models that do not suit platform models like new blockchain initiatives.
The service that is brought to you often through a platform service goes against the value that a decentralized solution (sometimes constructed on top of a blockchain) creates. A service provider implies a central authority (or more kindly, a central creator of value), exactly the opposite of what a decentralized application like dApp aims to solve. The holy grail of (platform) service-based companies is often seen in the form of an Initial Public Offering (IPO), which also implies a centralized creator of value. The new thing for decentralized businesses is to offer an Initial Coin Offering(ICO), which is suited for decentralized businesses. This has a tremendous impact on how you structure your products and services, too.
Good Software == Good Business
But when does this become an important topic? One argument of those who disagree with my views is that the first thing you need to do is focus on the problem you are solving. And that this should have the sole focus of what you do when it concerns a new business. Well, that might be true when kicking off, but in a matter of weeks, the actual product or service starts to play a role. It is not mutually exclusive. If you have a bad business idea but great software, it is easy to pivot, but a great business idea without software is not only useless, it also shows the lack of knowing how to execute.
Therefore I would argue that the software equals your business. Something which is significantly different from pre-software businesses where the recreation of value cost a significant investment relative to the software business.
What is Next?
All the things mentioned above are now status-quo when it comes to building new digital products and services, if you know how it works, you can see it everywhere. For me, however, the question becomes interesting when we think about what this will evolve into next.
The methods that we use to build software products and services are becoming standard processes and fixed methodologies and because of physical products being so deeply intertwined with digital asset (id est, software) we see that technology companies are moving into other domains, for example, construction and real estate. Those domains look completely different at first sight, while in fact they are physical processes that can be manipulated and disrupted in a similar fashion like it is done with digital assets.
If you are working on a new digital product or service, it will be important to decide early on how you want to offer it to the world. Are you using a more traditional approach, a service-oriented (XaaS) approach, or something modern like a decentralized approach? The moral of the story is that it has a big impact early on. Taking a wrong turn along the way might have a tremendous impact on your business, but then again, that’s also the fun, isn’t it?