Massimo Pezzini on the composable future of enterprise IT
- After 25 years as Gartner's top analyst on enterprise integration and automation, Massimo Pezzini has intriguing views on current market trends towards composable applications in enterprise IT.
End-to-end integration, effortless automation and composable applications have been the goals of IT for as long as I can remember. Are they finally within reach? Someone who's better placed than most to answer that question is Massimo Pezzini, who last year retired as a Gartner VP Distinguished Analyst after 25 years at the firm as a leading authority on enterprise integration and automation technologies. His answer is an emphatic yes, with huge implications for the future of enterprise IT. He says:
The application portfolio of a company in five years from now is going to look much more different than it is today in terms of the architecture — more building blocks composed together and less and less of these gigantic application suites, which are super-rich in functionality, but also very inflexible, very hard to deal with.
I caught up with Pezzini earlier this year, shortly after enterprise automation vendor Workato recruited him as an advisor on enterprise futures. Having known him since the 1990s, I knew we'd have a fascinating conversation, and discovered that we share similar views on quite a few of the emerging trends in enterprise IT, including the inevitability of cloud computing, the rise of what I've been calling co-code as IT and business teams work more closely to improve automation, and the transition towards composable applications within what I think of as Tierless Architecture. Adopting these emerging approaches is becoming a business imperative, he argues, that will soon become as self-evident as the need for an email system. He elaborates:
Having a sound integration/automation strategy is a business weapon ...
Today, we have seen organizations building the business case for integration/automation, how much money they are going to save, blah, blah, blah, time to value, etc. I believe that in maybe five years from now, these [business case arguments] will be gone.
Integration/automation infrastructure will be considered an essential business tool without which you cannot be in business ... because the ability to quickly compose and recompose your business processes — to deliver a new product to customers, to introduce a new service, to comply to a new regulation — is going to be vital. And without a sound integration/automation strategy, it's simply impossible to do this kind of thing.
Rapid growth in iPaaS, API management and RPA
This trend is already reflected in market dynamics, which show a clear trend of investment moving away from older integration and automation tools such as B2B gateways, application integration suites and so on, while newer tools that support this more composable approach — in particular integration-Platform-as-a-Service (iPaaS), API management and Robotic Process Automation (RPA) — are growing fast. With the market as a whole expanding by more than 30% annually, the growth in these newer tools is even faster. That's driven by three factors, Pezzini explains. First of all there's the move to the cloud, and SaaS in particular, with the public cloud market set to grow from an already massive $125 billion in 2020 to $280 billion in 2025, according to Gartner forecasts. He comments:
As more and more of the application portfolio organization moves to the cloud, there is a clear need for an automation and integration platform which also runs in the cloud.
Cloud also has the advantage of being able to deliver innovation faster, he notes, which makes it particularly attractive at a time when companies are seeking rapid competitive advantage. Secondly, these newer tools encompass more functionality. He explains:
From a use cases point of view, they can address a little bit of data integration, a little bit of application integration, a little bit of API, a little bit of BPM, a little bit of B2B, etc, etc ... Because those products are functionally richer, they have a larger functional footprint, and therefore they are stealing business away from the more specialized, classic software.
Finally, the ease of use is dramatically simpler than for traditional BPM and integration tools, which shortens the learning curve for these newer platforms. He comments:
You don't need to hire an army of PhDs in computer science to use these tools, which means that the barrier to entry dropped down dramatically ...
The traditional integration software, you had an integration competency centre of 10 people, and they were the only people in the company that could use this technology. If you throw in Workato or Jitterbit or one of these tools, a lot of people can use this technology.
CIOs become facilitators
Pezzini argues that CIOs need to get out in front of the resulting "democratization of integration and automation" by becoming facilitators of this new wave of development led by Line of Business (LoB) managers. He explains:
I think that, going forward, more and more the role of the CIO, the model for the IT department will be to act as a service provider, as a facilitator, as an enabler. In other words, I think it is — in many cases, not always, but in many cases — it is appropriate for the LoBs and the departments and the business units to build their own solutions. But the IT department should provide them the enabling technologies, as a service. 'Okay, Mr VP of Sales. You want to build this e-commerce application? Great, you can do it. Here is a set of enabling technologies that you want to use to make sure that security is appropriate, that the compliance is appropriate, that the integration is appropriate, etc.'
I, as the central IT, am going to provide these capabilities as a service. I'm going to monitor and manage what's going on, co-ordinate the entire thing, but you are responsible, you are in charge, you develop, I'm here to help you. I'm here to provide consulting and support and training and mentoring, etc. You do whatever you want, you do it on my shared platform so we optimize cost, etc. And in this way, of course, the CIO has some degree of control over what's going on.
He points to analytics and BI having followed a similar path with the move away from dashboards and reports being built centrally by BI competency centers, towards those competency centers supporting more user-friendly reporting tools where people organize their own data analysis. A related trend is the emergence of what Gartner calls 'fusion teams', in which business and IT people work collaboratively on application development. He now sees similar trends taking shape in integration and automation, as he explains:
Central IT [acts] as an integration strategy team, which provides the tools and the capabilities. But the line of businesses and subsidiaries and application teams etc, they help themselves, fundamentally under the supervision and governance of the central IT. This is I believe, how we can compromise between the extreme flexibility and creativity of the full democratized environment, and the rigor and compliance and governance of a fully centralized environment.
Enterprise apps become composable
Meanwhile, established enterprise application vendors have to adapt to this new IT landscape, in which the old monolithic suites are being replaced by interconnected ecosystems of multiple applications. Pezzini elaborates:
In the old days you would buy SAP ECC, and in the package you have CRM, finance, procurement, expense management, HCM, you name it, manufacturing — everything in one box. Now, from SAP you buy S/4 HANA, for the core administrative processes, you buy SuccessFactors for HCM, you buy Concur for expense management, etc. They are disrupting themselves, to a certain extent, by breaking their monolithic big systems into smaller chunks, because of Software-as-a-Service. They had to do this, because Software-as-a-Service was eating their cake.
He believes the process will continue to a logical conclusion, in which enterprise applications become an ecosystem of composable building blocks, as he explains:
At some point, I believe that the application landscape of an organization will look like a broad set of elementary business components — accounting, payables, receivables, tax calculation, what have you — possibly coming from different vendors. The end-to-end process, the end-to-end application, will be built by these fusion teams, using orchestration tools. Teams will use these tools to aggregate and compose together these component building blocks at the backend, and rearranging them and shaping them in the way which fits with the company's business needs ...
Hence the importance and the relevance of low-code, no-code tools and integration/automation platforms, such as Workato and the others, which fundamentally puts in the hands of the business users the ability to create their own business processes in the way they want, by essentially assembling and aggregating together pre-existing component parts.
Advantage in this new world comes not so much from the individual components that vendors provide, but in how the enterprise connects them all together. He explains:
What's going to be the difference between my company and your company is not the components, because we can all buy the same components. But the way in which we aggregate and put together those components is going to make the difference.
Connecting the dots
Connecting up data across the organization and making it visible is particularly important, as he explains:
At the end of the day, digital is about the ability to capture business events, such as a customer wanting to buy a product, or a supplier not delivering the product, or the spare parts, as expected ...
An API is a way for my core business processes to know that customer XYZ wants to buy product ABC right now. Usually those data end up in a data lake or data warehouse. But what if, while those data are moving around through the integration/automation infrastructure, what if we plug in some real time analytics? We can build real-time business dashboards for the CEO of the company or the head of sales or regional manager, so he or she can see in real time how the pipeline is moving, or how the supply chain is performing. In this way, we can also detect situations which require our attention, for example, a disruption in supply chain ...
So the idea is to leverage the integration/infrastructure not only to move data around, but also to utilize this data, leverage this data, to surface these situations which require our attention. But this is only possible if you have a sound, well-defined, well-managed, well-governed integration and automation infrastructure.
Evangelists for this new approach have a challenge in explaining how it works to business leaders. Pezzini says:
Having strategic investments of integration/automation technology enables you to achieve these business goals. But it is not easy for the business side of the house to understand this. I used to say that the tragedy of automation/integration is that you cannot make a demo — there is nothing to be seen!
The benefits, however, in terms of efficiency, agility, user experience, innovation and data-driven decision-making are evident once an organization begins to go down this path. He sums up:
Invest strategically in integration and automation, because in this way, you can save money, you can be more agile, you can be more innovative, you can improve the experience for your customers, and you can have a better understanding of what's going on in your business.
I was surprised to find in this conversation how closely our views aligned. I think it's particularly interesting to read this interview in the light of last month's MACH One conference, which focused on similar trends in the field of online commerce and content. That's partly why I held off publishing this interview until now, because I believe these two viewpoints are two sides of the same coin. At the same time that builders of brand new digital application infrastructures have adopted a composable architecture, so too there's a new wave of integration and automation specialists that are shifting existing enterprise IT into a more composable idiom, and I'm confident the two are about to converge. Alongside of that convergence, a new approach to application development that brings IT and business people closer together is also emerging in both of these fields and again it's all part of the same trend. My term for this trend is Tierless Architecture, and drawing on earlier writings along with these more recent articles, I'm now ready to write up the next article in my extended series on Frictionless enterprise to put all of this in context. Watch this space.