After a quarter-century of SaaS, the next 25 years belong to XaaS

Software-as-a-Service (SaaS) — the use of software via the Internet, rather than locally on a standalone computer — has come to dominate computing over the past quarter-century, along with Infrastructure-as-a-Service (IaaS) and other siblings. That history holds lessons for the next 25 years, as the digitally connected business model its providers pioneered reaches out beyond computing to touch every industry. Products, services, experiences are all being swept up in the phenomenon of Everything-as-a-Service, or XaaS (pronounced 'x-aas'). Here's what this means.

Many of the points I'm going to set out here came up in a discussion I took part in last week with Jeff Kaplan of THINKstrategies and hosted by Deal Architect's Vinnie Mirchandani as part of his Burning Platform series of interviews. In a two-part discussion, we looked back on 25 years of 'as a service', and forward to the trends shaping its future. Jeff and I were both prominent as industry analysts at the heart of the early years of SaaS and cloud, while Vinnie has more than three decades' experience as a leading analyst and author writing about enterprise computing and innovation.

XaaS is already taking over

One of the lessons of history is that the new model starts taking over before people realize what's going on — no one is looking for it, so it stays hidden in plain sight. Back in the early 2000s, when people were still debating whether enterprises would ever trust their serious business endeavors to Internet-based providers, this was already happening before their very eyes. Google was already growing a multi-billion-dollar business selling Internet advertising-as-a-service in the form of its AdWords offering, while eBay and Amazon grew multi-billion-dollar businesses with platforms that enabled companies to sell online.

Today, the same is happening with XaaS. The music and entertainment industry has already largely switched to a digitally connected XaaS model through the likes of Spotify, Netflix and their rivals. Publishers are moving from an ad-supported business model to subscriptions. Even retailers are turning shoppers into engaged subscribers with the likes of Amazon Prime. XaaS is already all around us — and it's not going away. Here's my take from the closing minutes of last week's discussion:

I think it's probably more prevalent than we already realize. And I don't think it's going to go backwards, and the reason for that is because of this digital connection. It just makes it more viable as a way of doing business than ever was possible in a disconnected world.

So, you know, it may go away if the Internet stops working. But until that moment, I think it's starting to become more and more mainstream.

It's all about the connection

So what exactly is XaaS? My take is that most of the commentary about XaaS, when people simply use the 'X' as a shorthand for all kinds of IT — Deloitte is a recent example — is that they're missing the most important element of the XaaS model. SaaS — along with IaaS, Platform-as-a-Service (PaaS) and all the rest — was transformative for the software and IT industry because it created a continuous digital connection between the provider and its customers. And that is why the XaaS model will transform every other industry, as every product, service and experience becomes digitally connected.

Most of the commentary about all the many varieties of XaaS in the IT industry ignores this aspect. It sees the impact of XaaS as a way to lower running costs, free up resources, speed implementation, move to a pay-as-you-go subscription, and get better access to up-to-date technology. All of this is true, but what the early SaaS providers discovered was that continuous digital connection also enables a three-stage virtuous cycle that is the secret sauce of the XaaS model:

• Engage — the customer is directly connected to the business every time they use the product or function.
• Monitor — the business can see how all of its customers are using its products or services, and proactively fix any problems.
• Improve — the direct connection allows the business to update its software or offer new capabilities.

Continuous improvement is paramount

Much of our discussion about future trends last week centered on this aspect of continuous improvement. Many businesses misinterpret the XaaS model as simply a matter of changing the payment model, much like leasing a car instead of buying it outright. But as Kaplan pointed out, "When you lease a car, you take it off the lot, and nothing really improves unless it's a Tesla." And why is Tesla the exception? Because it's a digitally connected product whose manufacturer remains engaged with its customers, monitors their usage of the product, and improves its functionality and performance over time by upgrades to its software. XaaS providers who don't offer this continuous improvement won't succeed in a competitive market. As Kaplan explains:

In this as-a-service world, you subscribe based upon whether there's an annual or multi-year agreement, and you expect that the SaaS vendor, if they're following the best practices in the marketplace, are adding more features and functions every quarter. So I think that that continuous improvement process is happening among the best companies in the industry.

We also discussed best and worst practice examples. I was strongly critical of providers that use intellectual property laws as a pretext for locking customers into their own proprietary and costly maintenance services, which I would class as an abuse of the XaaS model and one that customers should be rightly be wary of.

On the positive side, we discussed the use of machine learning to analyze data aggregated across an XaaS provider's customer base and provide benchmarks back to customers, showing where they can improve their own performance. One provider that's been doing a lot of work in this area is business spend vendor Coupa, whose CEO Rob Bernshteyn makes a persuasive case for what he calls this community intelligence approach. My observation is that the use of machine learning and AI gives a huge boost to the value of this approach and thus makes the payback for customers to share their data in this way much more attractive. I think this means we'll see this becoming much more prevalent and it will raise the competitive bar among providers.

How do you measure customer success?

One of the most important considerations however is what to monitor and how to measure improvements? The XaaS model moves the focus to outcomes, which is different from simply looking at customer satisfaction or customer experience. Traditionally, these measures focus on the customer's feelings about a transactional event such as the purchase, the installation, or a service call. In the XaaS world, the emphasis is on the substance of the customer's ultimate and ongoing success in achieving their goals as a result of using the service.

The state-of-the-art here is still at a very early stage of evolution. Currently, those providers who take this seriously are still largely focused on the customer's success in using the product. They look at measures such as adoption, what functionality is being used, and how quickly an action is completed. That's important, and doing this well already provides a competitive edge simply because most providers aren't even trying to do this at the moment. But my expectation is that the next phase of evolution will see businesses focusing much more on customer outcomes, with a corresponding shift in how they interpret and deliver the engage-monitor-improve cycle:

• Engage — what does success mean for your customers?
• Monitor — are they experiencing success?
• Improve — how can your business, product or service help them be even more successful?

One question that Mirchandani posed right at the beginning of the futures segment of our discussion was, why are there so few examples of providers pricing their wares on an outcome basis? My answer is that this is one of those cases where we can learn from the history of SaaS. The simple explanation is that the tools don't exist to measure outcomes properly. If you look back to the early days of SaaS, there were no software tools to manage subscription billing — providers had to manage everything on spreadsheets and custom-built apps. Consequently, their subscription offerings were very limited and it was hard to make even simple adjustments mid-term, such as upgrading from one level to another halfway through the year. This is the main reason why the per-user per-month subscription billed annually is still so prevalent across the industry.

Nowadays it's easy for providers to sign up to an off-the-shelf subscription billing service, with much more flexibility to experiment with different offerings. But outcomes-based pricing is as poorly supported today as subscription billing was back in those early days. Anyone who wants to offer it has to build their own system to track the various metrics and calculate the appropriate charges, which significantly limits the scope for experimentation.

The other brake on development is the lack of any consensus on what metrics and outcomes to track so that the resulting charges are accepted as fair by both sides. At the moment, this has to be determined for each provider and its customers on a case-by-case basis. It also depends on customers being comfortable with an outcomes-based billing model, which in a B2B environment may depend on their adoption of the XaaS model with their own customers. Many of the obstacles therefore are systemic and will take time to dismantle.

It's hard to change old habits

What stands out as the biggest lesson from the history of SaaS is how hard it is to go against the grain of mainstream behavior. The lack of tools and established practices for subscription billing is one example — in fact SaaS faced huge technology challenges across a whole range of capabilities that today we take for granted.

Another factor we discussed last week was that in the early days of SaaS, business technology buyers had been locked into the habit of buying software as an asset. This meant their spend came out of a capital expenditure (CapEx) budget, whereas paying a monthly subscription comes out of day-to-day operational expenses (OpEx). This drove the takeup of annual subscription terms, often signed up on two-, three- or five-year contracts and paid as a single lump sum out of the CapEx budget. It was a long time before buyers adjusted to a subscription model, and even today many enterprise buyers balk at open-ended pay-as-you-go contracts for variable services such as IaaS because they prefer to have a fixed commitment in their budget.

Now that SaaS and cloud have become the mainstream choice in enterprise IT, it's easy to forget how much of a struggle the pioneers faced to establish a model that was seen as the time as a passing fad that would never take hold. The mainstream locks everyone into ways of behaving and it takes courage and determination to think outside that box.

A new wave of programmable platforms

Another question from Mirchandani asked how SaaS vendors can satisfy the varied demands of customers across industries and geographies. My response was that there's a new more programmable architecture emerging, which provides a great deal more choice to customers in how they build up the functionality their organization needs — to "pick and choose the bits that you need to to achieve what you want to do."

This much more composable, tierless architecture is based on APIs and gives customers the freedom to mix together best-of-breed components from XaaS providers with their own custom elements developed in-house (though typically still deployed in the cloud). As Jeff mentioned, Chris Barbin, who co-founded Appirio, one of the pioneering Systems Integrators (SIs) in the early days of SaaS, has recently launched a venture fund to invest in a new generation of SIs to serve this emerging segment.

There's often a tension going on between vendors, who want to provide as much as possible of what their customers are using, and buyers, who don't want to be at the mercy of a dominant supplier and therefore seek to keep their options open. Here's a case where many leading SaaS vendors are themselves now part of the mainstream, and they'll need to adapt if they are to retain their hold on the market.

Look to new entrants for innovation

Another important lesson from history is that the tight grip of mainstream patterns and mindsets means that you shouldn't look for innovation to come from the big, established players. We saw this again and again throughout the history of Saas and cloud. However much the incumbent vendors thought they were innovating, they were constantly pulled back towards the old ways of doing things. Even when they came out with SaaS offerings, in the early days it was just their existing software deployed unchanged into the cloud, a phenomenon that at the time I termed SoSaaS — Same old Software, as a Service. It's taken more than a decade for the likes of Microsoft and Oracle to finally figure out how to do cloud and SaaS — and they still have to work hard to keep up with the smaller, younger innovators.

By the same token, don't judge success by the same criteria as the incumbents. SaaS and cloud grew in hidden sight because the likes of Salesforce, Google and Amazon were not displacing existing forms of enterprise application but inventing new categories. It's about finding new patterns enabled by the technology that weren't possible before, and which offer new value in today's digitally connected world. As I said to Mirchandani last week, retailers shouldn't be looking for direct replacements for their on-premise merchandising software. They need something different in an omni-channel Vaccine Economy, where customers are no longer making their buying decisions in-store. Or look at payroll. Instead of providing a service focused on employers, payroll software should be oriented around the needs of individuals who have side-businesses and gig jobs alongside their main employment and need a service that aggregates their payments and tax calculations.

Later in the conversation, we talked about the failure to tempt farmers into an XaaS model for agricultural equipment. I narrowly missed my chance during the conversation to repurpose my acronym for this new world, so let me conclude by taking that opportunity here.

My take

There is no sustainable market for vendors who simply package up their Same old 'X', as a Service (SoXaaS). Adding a digital connection and a subscription model isn't enough to create a true XaaS proposition. It requires real engagement with customers to understand the outcomes they're aiming to achieve, monitoring how your product, service or experience contributes to that, and a real commitment to continuously improve the offering so that it supports their success in those goals. This is the secret sauce of XaaS and it defines the next 25 years of digitally connected business.

Last week's two-part discussion can be viewed on YouTube via these links — 25 years of 'as a service' and The future of 'as a service'.