Modern Groundhog Day – When will modernization end?

This article was first published on Medium, read the article here.

 

On February 2, 2021, the Groundhog Day, Punxsutawney Phil emerged from his snowy burrow to predict six more weeks of winter! “We have all passed through the darkness of night but now see hope in morning’s bright light. But now when I turn to see, there’s a perfect shadow cast of me. Six more weeks of winter there will be.” narrated Phil’s top-hatted handler.

If you are an executive responsible for Data & Analytics Modernization in your firm, you may well be living what Phil Connor, played by Bill Murray, lived through in the 1993 classic “Groundhog Day”–a time loop. Except, you find yourself in what I call “perpetual modernization cycles”– a result of exponential growth of different forms of technological progress over the past two decades–something suggested 20 years ago by American futurist, Ray Kurzweil, in his renowned essay, “The Law of Accelerating Returns.”

The Modern Groundhog Day – Start of Perpetual Modernization Cycles

As we embark on a new decade, I can’t help but think about what IT executives will have to deal with in the coming years. We already face relentless pressures to modernize. Business is creating more demand for speed, agility, and capabilities; while, IT is facing supply pressures due to changing technology, ever-growing data sources, and security threats. According to Gartner, by 2025, 90% of current applications will still be in use, and most will continue to receive insufficient modernization investment.

To address modernization, companies are using approaches that look something like Rehash, Rehost, Rebuild, and Replace using an effort and value trade-off. There is no one-size-fits-all approach and you are likely trying multiple options, like what Phil tried, using a combination of technology, functionality, and architecture adaptations to deal with your situation. But even before you finish your current project, you are already facing pressures either from the demand- or supply-side. Welcome to the Modern Groundhog Day!

During my 20-year experience supporting enterprise customers in the public and private sectors, I have lived through multiple cycles myself. For example, at the Department of Health and Human Services, I supported multiple versions of web-based systems used to collect, report, and analyze–clinical, performance, and administrative data from over 13,000 health center service delivery sites providing primary care to over 25M patients nationwide. At Infocepts, we are modernizing data-driven capabilities for multiple retail, media, data syndication, financial, and health companies across the globe. Whether it is preparing customers for the Olympics, or liberating them from legacy monolithic BI platforms, or using AI-driven insights to connect retail businesses with consumers–we are constantly driving modernization initiatives.

Like Phil Connor, we evaluate each experience and apply our learnings to future endeavors improving outcomes for our customers. Here I share three learnings and ideas to deal with the Modern Groundhog Day.

Three Implications of the Perpetual Modernization Cycles

The scope of modernization programs is becoming broader. Companies are no longer looking at “lift and shift” approaches to upgrade technologies, instead, becoming more mindful of revisiting end-to-end business needs to deliver the intended benefits such as improving customer experiences, reducing TCO, or improving operational effectiveness. But that means more decision points along the way, often beyond the IT organization, and thereby more time to execute. Agile techniques and modern technologies such as cloud and low-code platforms help with reducing the cycles times and time-to-value, but large-scale legacy modernization programs often span across multiple calendar years, despite what vendors want you to believe.

• 1. Longer cycles mean business keeps waiting for intended benefits

  Proponents of agile techniques–and, I am one of them–often advise clients to focus on some variation of “minimum operating capability.” After the initial discussion on the scoping and prioritization, the program management leadership focus on the question of, “What is the minimum capability necessary in production to see value?”, and allocate resources to get to that. But that isn’t the finish line and success shouldn’t be measured based on that metric alone.

  An Infocepts data syndication customer operates large-scale data factories; they are in the process of modernizing them–a few processes at a time. But the consumers of these factories residing in distributed business units are facing massive operational pressures to cut their annual operating budgets. The challenge is that the processes that haven’t been modernized continue to drain resources and perhaps the program management leadership is oblivious to that.

  So, the key takeaway here is to measure executive success of modernization programs based on the initial time-to-value, incremental cycle times, and total time-to-value with corresponding TCO numbers.

• 2. Organizations are accumulating redundant capabilities faster than they can retire their legacy ones

  To take advantage of technology evolution, companies are adopting new systems for engaging their customers and/or new systems for innovation while maintaining their systems of record. In a rapidly shifting market, vendors are offering niche capabilities to create differentiation to acquire customers. This is particularly true in the analytics landscape where platforms such as Tableau penetrated from the front-end visualization layer, closer to business users, and then organically matured toward established enterprise platforms such as MicroStrategy and Business Objects. The same is true now with products such as ThoughtSpot that are focused on search-based analytics, continuing to bridge the gap between the end-user and insights. This trend coupled with inorganic changes to customer data & analytics landscape due to internal restructuring or M&A results in multiple redundant capabilities.

  An Infocepts media customer ended up with six enterprise capabilities to include MicroStrategy, Tableau, Qlik, Power BI, Business Objects, and Looker! Clearly, they needed to eliminate redundancy and save costs. Solving such a problem requires a few months for the rationalization, then migration, and ultimately timely decommissioning of redundant capabilities. It took our team 12 months to cut down the number of systems from 6 to 3 and save them $750K in annual costs.

  The key takeaway from this is to maintain up-to-date information on current tools to support future rationalization efforts and tie decommissioning to project success. Maintain basic information such as the core purpose of the tool, usage data, financial data, and intangibles such as customer expectations to speed up rationalization efforts and to inform critical cutover milestones based on software renewal dates.

• 3. Leaders (and teams) are grappling with multiple responsibilities

  Regardless of how you are organized, IT executives typically deal with responsibilities to include strategy, experimentation, development, operations, and adoption. For any modernization project, you must address these elements–how you choose to do may vary. Factors that play into how leadership and teams are organized include the need for legacy knowledge, aspirations and interests of leaders and team members, and long-term vendor agreements and relationships. But you can see how quickly the number of responsibilities multiply with the result that teams start to locally optimize for their available capacity and the overall velocity suffers.

  For example, data migration requires knowledge of business rules prevalent in the legacy system; teams want to work on newer technologies and not be boxed into maintaining legacy tools; a leader may be interested in demonstrating results for their growth; and, companies may be locked up into master service agreements with global system integrators to gain preferred rates with penalty clauses for deviating! These dynamics hamper modernization success and the longer the cycles last, the more complex they become. The result is that customers pay more for their modernization efforts than initially forecasted.

  The key takeaway here is to ensure that companies take an end-to-end approach for modernization outcomes and establish cross-functional leadership teams to devise and execute the strategy to minimize cognitive biases from influencing enterprise roadmaps. Modernization is more than a technology or capability upgrade.

Three Ideas to Deal with Perpetual Modernization

As companies deal with multiple modernization cycles, I anticipate them codifying their holistic approach in an Enterprise Modernization Life Cycle (EMLC) much like the Software Development Life Cycle (SDLC). The key is to connect the EMLC to their enterprise roadmap that advances them along their pursued business strategy. If you do these two things, you’d be able to find comfort in its predictability and take advantage much like how Phil made use of each of his new days once he found his purpose!

• 1. Don’t reinvent the wheel

  With the proliferation of cloud-based “services”, it is becoming harder to quickly put them together to construct enterprise capabilities. This is in part due to the paradox of choices and our desire to make the right choice without negative repercussions. Some vendors are making it easier for businesses by offering “Anything-as-a-Service.” But if you are responsible to advance your enterprise roadmap you can look for building blocks in the form of “turnkey projects” or “reusable capabilities” and compose your bespoke solutions rapidly. In this case you are counting on the experiences and learnings from other parts of your organization or from the market to reduce your time, costs, and risks with success.

  Consider a situation where you want to find an alternative to your legacy analytics enterprise platform and move to a modern one. This is likely not a very “common” project for your team. In such situations, you may look at experts that do turnkey migration with precision enabling you to focus on making critical choices, supporting current operations, and preparing for the cutover. As an example, Infocepts offers a BI migration solution that consists of a cross-functional team with a proprietary methodology and tool refined over successful execution of 100s of such projects to guarantee your success. Similarly, we have a solution to move workloads from legacy data warehouses such as Netezza to Snowflake within a few weeks.

  You may find such solutions within your own organization as well. For example, at one of our customers, teams from different business units are sharing foundational capabilities for a CI/CD DataOps pipeline so they can augment it, rather than starting from scratch.

• 2. Get your teams out of the DIY mindset

  You should encourage your team to think outside the box when making progress toward modernization initiatives. This is to overcome situations where the same team is signing up for several capabilities and then prioritizing them locally. Teams with highly capable technical talent often think they can learn new technology quickly and then implement at scale. Or like mentioned earlier, some companies may be constrained by their master service agreements with penalty clauses. Regardless of your situation, you should evaluate your team’s current capacity and focus, and what you need to accomplish toward your modernization, and determine how to allocate responsibilities creatively.

  For example, one of our customers wanted to free up the capacity in their operational team toward modernization. They engaged Infocepts to do a turnkey RPA automation using UiPath to reduce both the time and effort to execute quality assurance lifecycles for their report factories. In just 3 months, we helped reduce the QA effort from 9 FTEs to 3 FTEs while increasing their coverage to 100%.

  At another customer, our team helps in experimentation while they keep the enterprise implementations closer to their core team. This could be a way forward to try out multiple options for newer capabilities such as Data Science adoption. The reason is that a single team or organization likely does not have the time or resources to exploit the breadth of scenarios during evaluation to help them make informed choices.

• 3. Consider alternative solutions

  The last idea prompts leaders to challenge their teams to explore alternative solutions to costly modernization endeavors. Cohesive agile teams sustained for long durations may get very productive, but there is a risk of them succumbing to groupthink. Let me illustrate with two examples.

  Consider a need within your organization to provide an analytics hub to make analytics assets from different capabilities more accessible to your users. If your team is used to building custom solutions or is currently supporting a custom portal, their default thinking would be to add features to their portal or build a new one. In such situations, you must challenge them using the build vs. buy rubric. For example, solutions such as Metric Insights and ZenOptics can provide hubs in a few weeks rather than going through a costly and risky development endeavor.

  Or consider a situation where you are looking to optimize long-running ETL jobs designed several years ago. Again, if your team is used to traditional warehousing approaches, they may Rebuild the ETL packages with automation. But such an approach may miss out on the Replace option using emerging technologies such as Incorta that enable direct connects to data sources with real-time aggregations and in memory transformations.

When Will Modernization End?

Kurzweil may well be on to something. We know for a fact now that the rate of technology is certainly changing rapidly. But for that rate of change to translate into accelerated returns for our businesses, we all need to think differently going forward. Rather than looking to modernize quickly, we need to find ways to remain modern. This can be done by adopting evolutionary design thinking, growth mindset, agile leadership, and reliance on proven reusable assets to charter our own course. Maybe, like Phil, we will wake up one fine day and find that we’ve transcended the modernization loops.