When it comes to technology industry trends, the pendulums swing back and forth with steady frequency. Numerous examples abound but two will suffice here: a) the shift from on-premesis to cloud and then a good bit back to edge again, and b) the move from green screens to native PC apps then to the browser. Now when it comes to the way IT is acquired and situated in the enterprise today IT, the pendulum is currently swinging quickly back to a highly-valued old position that will very likely separate leading IT organizations from the also-rans going forward.
Years ago, when IT was much less pervasive, corporate uptake of technology was more varied, and the giant enterprise suites hadn't yet become the all-consuming platforms they've become, it was a simpler time. Back then the CIO put all the strategically differentiating IT into one bucket, and all the commodity IT that didn't move the needle much for the business into another. Then IT took the applications and systems in the differentiating bucket -- the ones that could break the business if done poorly, or would make/keep the business an industry-leader if done right -- were carefully selected, acquired, and customized to fit the special way the business carried out that function, integrated with key systems and then brought online.
Great IT Helps Organizations Compete Better in the Market More Than Ever
As history is such a good teacher, particularly in the IT world, it's easy to look back to see that there were several problems that emerged from this approach.
First, integration was often the most time-consuming and expensive of all of these processes. The best IT systems don't create new data silos when that information already exists in other systems. But for most of the history of IT, integration was hard. It often delayed or even put the whole project at risk entirely through intractable integration issues discovered too late in the game. It wasn't until quite recently that lightweight integration approaches including open APIs, microservices, and now the giant new vendor graph APIs such as Microsoft Graph and SAP Graph emerged, which when combined with integration platform-as-a-service (iPaaS) like Dell Boomi, Jitterbit, Zapier, and Integromat (to name just a few) that made integration an order of magnitude less costly and time-consuming.
The second problem was IT customization. In the days before cloud, vendors allowed on-premises IT systems to be customized because that's what CIOs wanted, for the differentiating reasons cited above. Though customization was often overdone, the argument itself was sound: Differentiating can really matter, especially around the systems that drive the core value proposition of the business. Yet the IT customizations of yesteryear were frequently stranded as the vendor upgraded the product, which often made the old customizations outdated or incompatible. This required expensive redevelopment and testing for each upgrade cycle. As we now know, it's much better to quickly iterate IT systems to keep up with the market and changing business needs. Yet customizations had the opposite effect, leaving most organizations trapped on rapidly aging versions of products. So customization as often abandoned. As much as it brought strategic value, it also required organizations to take on too much risk and cost, at least with the state of technology back then.
Second Wave DT Requires Difficult-to-Achieve High Velocity Change
Then cloud arrived, along with software-as-a-service (SaaS.) Now IT had competition, and real competition it was too. Apps could be quickly procured and deployed in weeks at lower price points, integrations were easier with standard, simple APIs via REST, and most vendors now have a growing directory of off-the-shelf integrations to make it easier and less costly. Even the entire data center could be made someone else's capital investment issue and maintenance headache.
But as organizations digitized everything more and more, it became clear that the most disruptive market differentiation was regularly demonstrated by cloud-native tech companies that were moving faster, innovating more, and creating more compelling digital offerings in the market. They were also doing this on more modern platforms, they maintained less techical debt, and provided new types of experiences that better appealed to businesses and consumers alike. We are now all familiar with what Netflix did to video rental companies, Amazon did to much of retail, what the iPhone has done to a dozen industries at once (media players, cameras, GPS devices, etc.)
Most businesses have discovered that they've digitized, but they haven't transformed, which take truly sustained and widespread effort on the technology front.
Now we're in a second age of digital transformation (DT), where every organization now has to be a fast-moving cloud-native tech company in order to compete. Unfortunately, most organizations have at least three primary issues that are holding them back on the technology side: a) Two types of technical debt, one in acquired infrastructure/apps and another in in-house enterprise architectures that are too old and out-dated to compete yet are also extremely costly to replace, b) no appealing migration path from this situation to a new and better future that doesn't have a ten year program plan behind it (something well outside the 2-3 year planning and relevancy window of just about any IT department), and c) a runaway IT landscape that today has organizations acquiring and using more apps at an apparently exponential rate.
If you could ask a CIO what their biggest challenge is today -- after the out-of-control, wildly unpredictable, and potentially career-ending threats of cybersecurity and ransomware -- it is this: IT has simply become too complex in its current form to guide and navigate as an overall portfolio. This is not a theory. It is borne out by recent data: As I explored recently on ZDNet on this subject, most CIOs now cite tech complexity as their #2 challenge overall.
Further complicating matters is that the major enterprise IT vendors have built out -- through both development and acquisition -- now massive enterprise suites, especially in ERP and in front office productivity -- which have individual offerings that tend to vary substantially in quality, maturity, and overall integration with the rest of the suite. If given an easy choice, CIO would like to mix and match best-of-breed parts from multiple vendors, but prior experience has shown it was too costly and risky.
CIOs also don't like to become beholden to too few vendors (who then hold the cards), for reasons of choice and having too many eggs in one basket. However, while they do like that the modules of enterprise suites do to work together a bit better and they only have one throat to choke for problems, it's increasingly becoming moot as more and more parts of the business punt locally and use a better tool for the job anyway.
What then can IT leaders do today to address all these factors?
Differentiation Comes Back with a Vengeance
Most CIOs would very much like to put the complexity genie back in the bottle. In fact, they'd like to make the IT landscape less complex for a whole host of reasons, a partial list of which incudes fewer data silos, more consistent user experiences, easier worker onboarding, lower support costs, less cognitive overload, more strategic differentiation, more business productivity, less vendor lock-in and an easier overall IT portfolio to manage and govern. Yet it's likely that this complexity will eventually grow beyond human ability.
Consequently, successful managing complexity, as I began to say a few years ago, is fast becoming an IT imperative, not a nice-to-have. Soon we will simply have too much homogenized IT that doesn't separate us meaningfully from the competition, but still accompanied by sheer data and experience sprawl that will make it very difficult to have workers learn and do their jobs effectively (other than perform whack-a-mole across many dozens of highly variable and poorly (or not-at-all) integrated apps.)
As I've argued recently, the starting point for addressing these challenge effectively requires a new and highly-related focus and mindset: We must become experience driven organizations. Where to start in this view? We must focus on the most strategic and differentiating areas of customer, employee, and even partner experience.
However, putting the IT complexity genie back in the bottle is going to require a sea change in a few key parts of the IT landscape. One that would ordinarily seem unlikely to happen on its own. Yet happen it has with the arrival of highly capable entries in two relatively new product categories and one still emerging one. These have been directly enabled by the widespread adoption of APIs by SaaS and cloud platforms vendors. Combined, this has created an emergent industry breakthrough that has returned mix-and-match power to the CIO, as well as a way to combine and streamline app experiences into simpler, easier to train on and support solutions. Business data can easily be accessed wherever it's needed, while individual features can be extracted and moved to the places where they best belong, with the swift assembly of pre-existing data and functionality into compelling new point experiences. This approach can also be scaled better, because it uses high leverage new conceptual models that are as as manageable as they highly accessible, so that mixing-and-matching can be a daily activity carried out by a much wider group of citizen developers, instead of just IT.
Indeed, this self-service model, one of the key new ways to tap into much larger resources with low cost (such as end user peer support), has now arrived for integration and application development as well.
Three Key Enablers of Best-of-Breed IT: The Product Categories
While there are other supporting technologies in best-of-breed IT, the principle categories that are most influential in moving IT into a new mix-and-match posture, yet and also delivers on the experience imperative for more streamlined, more-focused, and situated IT at high speed and scale are:
- Integration Platform-as-a-Service (iPaaS). The industrialization of integration has arrived with platforms that have combine off-the-shelf integration catalogs with point-and-click "wiring together" and testing. Performance management, operations capabilities, and governance are also common in iPaaS. The best iPaaS even makes it possible for power end-users to integrate much of their data sources, with the aforementioned Zapier and Integromat being prime examples. With iPaaS, most organizations can quickly API-enable and integrate most legacy IT.
- Low-code/no-code solutions. While spreadsheets have long been the low code tool of choice for the typical non-IT worker, they are entirely inadequate as a low-code model. The good news is that this has begun to change in a big way with the advent of solutions that bring application development-style sensibilities and features (testing and version control, for example.) Our regularly updated Low-Code ShortList which I maintain with my colleague Holger Mueller is a good example of how a cottage industry has grown in a major tech industry in its own right. Many analyses have shown that there will never be enough professional coders to digitize everything that needs to be digitized in businesses today. Nor does there need to be. Making it easier for workers to turn their tech dreams into reality is now the next step in IT. That is, as long as the necessary guardrails around data security, privacy, and regulatory concerns are incorporated into the model. Many of these platforms now directly address this need.
- AI-based app generation. On the cutting edge, and without any clear category name yet, is what will likely prove to be the next major breakthrough in scalable IT differentiation. This is the creation of applications by using a capabile AI-based solution that can take high-level requirements in natural language, determine what the solution needs to do, then design, develop, integrate, and test it, and finally provide the completed app to the user for acceptance. While there are virtually no products in the commercial space currently available for this (yet), the impressible results of OpenAI's GPT-3 has shown that this breakthrough is not only possible but likely inevitable. The power this will confer on the average IT user is difficult to understate. It's is akin to literally verbally articulating the need for an app to solve a particular problem, and then having an AI agent immediately develop it.
Important Note: iPaaS and low-code/no-code are often used together to form what I refer to as a multicloud experience integration stack. This is an end-to-end strategic platform solution that helps deliver on best-of-breed IT assembly line with an experience-centric focus. Please see my exploration of this important new category of IT solution and the key players.
A New Democratizing Craft Model for IT, But Without the Old Baggage
How then can the CIO begin moving back to best-of-breed IT? First, they will need to create a better API-level foundation, or best-of-breed is relatively inaccessible to them if on-premesis IT resources need to be glued together or mixed-and-matched (which is highly likely in any larger organization.) That's because while their cloud and SaaS vendors are largely providing the APIs for IT already, most legacy systems still lack them. Fortunately, solutions like Mulesoft and TIBCO Cloud Mashery can great help added the needed legacy APIs in a full lifecycle framework. Second is the mindset that IT must be made far more malleable and easily-reshaped in the middle-tier and in experience-delivery, and that those that can and should engage in this activity aren't always in IT.
Most IT organizations do very much want to go faster and do better with digital, and now they really can. Employee experiences can be created for each step in the employee lifecycle out of the whole cloth of existing IT. New customer experiences can be rapidly created that are highly contextual and personalized, using everything the organization knows about them to improve their digital journey. But IT departments will have to unlearn a lot of their own predispositions and pick up some new talents along the way (especially low code education and application lifecycle management on the edge.)
That all this can be done by most organizations, there is little question. In fact it has been happening in the margins of many IT departmenst for years already. Now it must move from the margins to the center of IT. Ultimately, there is a clear and substantial business case in being more intentional and programmatic about creating a faster, nimbler, more malleable best-of-breed IT department. It's now up to the CIO to drive this agenda.
My research on this topic: