Click to learn more about author Piet Loubser.
Whether you are a close or casual observer of the technology landscape, it is clear that the advent of cloud computing, software as a service, big data, and near ubiquitous broadband connectivity are changing the way organizations apply, build, and use technology in a profound way. The traditional multi-month process to complete a three-year hardware capacity needs analysis, then to procuring the hardware, getting it commissioned and, finally deploying the desired application. That now can be completed in a few minutes using your PC and swiping your credit card. In fact, “Capex vs. Opex” spend categories, “agile”, and “faster time to value” are the buzz words of every sales conversation today.
While the benefits we are achieving from these faster and cheaper (at least initially) options are pretty solid, we are beginning to experience some challenges, as these agile approaches put tremendous stress on every aspect of IT – people, processes and technology. It is in this context that Gartner has proposed a bimodal IT operating model framework for how information technology should be managed in an enterprise. Mode 1 IT is similar to the traditional way of managing and running IT, which is to understand the requirements and use systems that do not change frequently. Mode 2 IT, on the other hand, relates to the exploratory, agile initiatives that go through fast iterations to either prove them out or fail and go away. Over the next year, expect to see the adoption of Mode 2 IT to accelerate, powered by platforms that will help manage this at enterprise scale.
The Yin and Yang of Mode 2 IT
As is the case in many advancements, there will be pros and cons to Mode 2 IT including:
- Improved agility and lower costs: As a marketing leader I have been fully embracing a lot of the Mode 2 approaches in selecting and deploying parts of my marketing technology stack. The benefits to signing up for a SaaS app like UberFlip for content management for the website or a new analytical tool to help with lead analysis in Marketo are plentiful. In most cases, it is easy to get up and running extremely fast, create the business impact as intended, and accomplish tasks at the speed of business.
Another benefit of the Mode 2 approach is that the cost to get started tends to be extremely low and can be paid monthly or quarterly. There are no lengthy cycles for correctly sizing, ordering, and commissioning the necessary hardware, and then implementing your software on top of that.
- Application and data integration challenges: While the functional benefits that led me to drive better content engagement on the web were plentiful with UberFlip, the challenge to integrate it with the rest of our technology stacks was not. Getting a single dashboard with all of the insights generated across my different apps became a bigger challenge with every agile solution that was acquired.
Scaling Mode 2 IT to the Enterprise
Speaking with analytics and technology leaders over the past few years, it has become increasingly clear that building a truly successful bimodal IT infrastructure and organization is more than just embracing agile and you are done. It is not just about turning everyone loose to try and do things as fast they can. Rather it requires a planned and managed approach to enable the entire enterprise to become agile. Over the next 12 months, expect Mode 2 IT adoption to accelerate with the following attributes:
- Increased adoption and deployment of self-service Data Management: Until now, self-service analytics has allowed everyone to pick their own tool of choice and then hack away at the data they receive or procure it in any manner possible. Individuals may feel that they are able to get things done quickly, but the reality is that, as a whole, the enterprise is not.
- Enhanced need for governance and compliance: Agile self-service processes are becoming and serving mission critical processes. For instance, a retail bank can massively accelerate stress testing by putting the power in the hands of the financial and risk analysts. However, the ability to manage and clearly understand end-to-end data lineage is of critical importance.
- Intelligent automation and operationalization of self-service workflows: With the growing need for user-built data flows, there is an increased requirement for them to be scheduled and automated by business analysts. The challenge, though, is how to intelligently detect dependencies across multiple projects and then automate these into single automated jobs that can be scheduled, restarted, and tracked.
- Growing consolidation into end-to-end analytical platforms: While the market has seen explosive growth in new analytical tools, including BI, AI, and other point apps, each of these applications runs into enterprise deployment challenges when access to clean, consolidated, and contextualized data is not readily available for the hundreds of analysts an organization wants to empower. The old adage that 80%of the job is about getting the data clean and ready quickly becomes reality here. Analytical platforms increasingly will need to manage the data all the way from its raw state through preparing and cleaning, and then ultimately getting it into the analytical or prediction engines to produce the key insights.
As we continue down this path of the consumerization of technology in general, whether it is compute, apps, data preparation, or analytics, it is becoming pretty clear that the line between Mode 1 and Mode 2 IT is not static. Nor is it the same for all organizations and industries. As you embark on your journey towards enterprise grade Mode 2 IT in 2020 or beyond, keep in mind that Mode 2 increasingly will encroach on the traditional IT-only domains.