iv. Pillar One: Strategy and Vision
Pillar One: Strategy and Vision
There’s an incredibly important transition in the broad information technology space that is often lost in the furor and excitement over generative AI.
You see, since IT time immemorial most chief information officers and those in similar roles have been called on by their organizations to essentially function as superintendents of utility companies. Their charge has been to keep the phones ringing, the emails sending and receiving, and to prevent data from leaking.
AI is upending this paradigm, even though many still don’t yet realize it. As AI and its related technologies become more pivotal to the success of an organization - see our earlier statistics regarding productivity and Investment - technology leaders are finding that they must transition from being superintendents of utility companies to being strategic leaders of the organizations they serve.
But crafting, executing, and making smart investments in scalable cloud and AI strategy is hard. Leading strategically - and empowering your people to implement the vision - can seem overwhelming.
Simply “wanting AI” doesn’t cut it. So, our Strategy and Vision pillar sets forth five dimensions which begins with vision, extends to creating the actionable roadmap and architecture necessary to actualize that vision, and finally establishes the programmatic elements necessary to drive that vision to fruition. These dimensions help organizations formulate and take action on their big ideas.
Executive Vision
We’ve tried in vain over the years to accommodate shortcuts demanded by various organizations with whom we’ve worked. Alas, we’ve reached the same conclusion each time: Technology adoption fails when not driven by executive vision. Adopting AI is simply too challenging for most organizations to do when absent of long-term vision supported from top-down. You simply must define the organizational direction of travel for AI at the CXO level.
This is the stuff of many, many business leadership books written over the years, so we don’t want to be too prescriptive here. Executive vision can take many forms, but the bottom line is that your executive vision for AI (or any technology) must frame everything that follows so that it is crystal clear why the organization is embracing this technology and what the organization collectively aspires to achieve from its adoption.
Figure 5 : A top-level executive vision for AI framing the aspirations for an organization embracing AI. This example includes building a digitally literate culture, creating a scalable and composable cloud ecosystem, extracting value from data safely, adopting a future-ready mindset and increasing AI knowledge and expertise.
We’ve helped many organizations craft their vision for AI. The anonymized aspirations below provide a great example of a top-level executive vision at a real-world enterprise firm.
Notice that our vision is aspirational, succinctly describing not just what we hope to achieve with artificial intelligence, but what we hope to be as an organization that has embraced artificial intelligence. Further, only two of our five aspirations explicitly mention AI at all. This is important: We often hear folks talk of AI as if it were a product, but it’s not a product at all. AI is quickly being woven through nearly every aspect of our work lives (and our lives in general), and it equally depends on the proper functioning of other domains including data, applications, technical governance, business process, digital culture, and the mission of the organization itself (“improving client engagement”, in the case of the example above).
Finally, a well-crafted executive vision ought to go beyond headline aspirations to describe what we call “targeted outcomes”, which is to say, to define the outcomes the organization hopes to achieve in actualizing its aspirations. Think of targeted outcomes as adding specificity to your aspirations, not necessarily hard, quantifiable specificity, but a clear articulation of what it means to (for example) “Extract increasing value from our data using responsible, safely leveraged artificial intelligence”:
The data platform offers a mastered single source of truth for the most mission critical data domains;
Data is addressable by AI and aggregated from different sources as part of our data platform;
AI is deployed consistently and with governance guardrails in place;
"Low-hanging fruit" (incremental) AI capabilities quickly deliver lower-risk capabilities to our colleagues;
We pursue a risk-sensitive portfolio of "differential AI" customized for the firm.
Whatever your executive vision, it is important to lead with it, to prioritize the AI investments that best align to it, and to evangelize it such that colleagues both in IT and the wider business understand the all-important “why”.
Actionable Roadmap
I Strategy without action is like the rule of law on a deserted island. Irrelevant, even to the birds
The trick to making strategy relevant is to pair it with an actionable roadmap, really the actions, activities, even full-blown projects that will be undertaken to actualize our aspirations and achieve our targeted outcomes.
There’s an old adage attributed to American General and later President Dwight D. Eisenhower that “plans are useless, but planning is indispensable”. Take it to heart. Firm roadmaps quickly grow obsolete even under stable conditions, and the only thing stable about the evolution of AI is its acceleration. An actionable roadmap for your AI strategy that runs more than 12 to 24 months into the future is far too long. We’re only able to achieve that level of durability by taking to heart our first principles:
AI strategy should offer immediate value to the organization beyond specific AI driven workloads because the nature and value of these workloads will remain unclear for some time. In other words, make investments in modern data platform technology that will pay dividends not just in AI but in analytics, business intelligence, search, etc.;
AI strategy must be flexible: able to absorb tomorrow what we don’t fully grasp today. It’s wise to plan 24 months in advance, but it is equally unwise to assume that you’ll not be regularly revising those plans as things evolve.
Figure 6: These five top-line priorities are representative samples similar to those that we see many organizations priorities as part of their early AI strategy.
Start by formulating up to five big priorities, inspired of course by your executive vision. If, for example, you have established five aspirations as part of your vision, try first to devise one major priority aligned with each aspiration. For example, referring to the executive vision shared earlier, we might establish the following topline priorities:
Then, add specificity to these priority buckets with 3-5 milestones that the organization will achieve in the next 18 (give or take) months. It’s helpful to break these down into three horizons of three to six months each, and be prepared to drastically rework the milestones in the third horizon given that they’re likely at least 12 months out.
Finally, keep in mind that you are likely to uncover specific actions or milestones you need to undertake simply by evaluating where the organization is in each of the twenty-five AI maturity dimensions outlined earlier. For example, if you assess early on that the organization is particularly immature in the dimensions of “AI Development Tools” and “Digital Literacy”, it’s wise to prioritize milestones that are likely to close those maturity gaps as part of your actionable roadmap. Finally, invest in your stakeholder relationships to ensure that your roadmap is mapped back to those stakeholders, clear feedback loops are in place, and updates are shared so that you bring colleagues on the proverbial journey.
Ecosystem Map
An ecosystem map is a high-level architectural diagram of an organization’s cloud ecosystem, and something that every organization ought to create at the start and continuously evolve as they progress on their AI journey.
The “map” metaphor is instructive here. It is used to distinguish an ecosystem map from the various forms of architectural diagrams, nearly all of which tend to include more technical minutiae than a typical ecosystem map. Whereas an architectural diagram provides specific parameters for specific technical solutions, an ecosystem map presents a higher-level, more visionary view of an organization’s cloud ecosystem.
To make an analogy to architecture in the physical world:
Solution architecture provides schematics - floor plan, dimensions, electrical wiring, ventilation, plumbing - from which a building is constructed;
Enterprise architecture provides plans for specific neighborhoods or systems such as a subway or electrical grid;
Ecosystem architecture and, by extension, an ecosystem map shows us the entire city.
Thinking of an organization’s cloud ecosystem as a city, we then conceptualize the next-level down component parts of the ecosystem as “neighborhoods” (we might have also called them “boroughs”). Cities the world over are pieced together this way: Downtown, Seaport, Southie, etc. in Boston; Greenwich, Soho, Canary Wharf, etc. in London (though you’re forgiven if you thought I was talking about New York until you got to “Canary Wharf”); Palermo, Recoleta, Puerto Madero, etc. in Buenos Aires; Norrmalm, Gamla Stan, Kungsholmen, etc. in Stockholm. The list goes on.
Each of these neighborhoods share the quality of dividing their city into smaller pieces, each often with their own distinct culture, aesthetic, or purpose.
Like cities, ecosystem maps are constantly evolving and changing. To prevent your ecosystem from becoming overcrowded, stagnant, or unable to meet the needs of its expanding 'population,' it's essential to revisit, revise, and adapt your Ecosystem Map on a regular basis.
The example above is one of our recent favorites produced in collaboration with a global enterprise organization as part of their AI and broader cloud strategy. Note that:
Specific AI and other data products are identified in the Data Distribution Neighborhood (bottom of the box on the far right);
The ecosystem map shows how data will flow from the organizations’ core business systems (including some yet to be identified, which is just fine) and application portfolio such that it hydrates various data distribution points, including but not limited to AI workloads;
Migration of legacy applications and legacy data warehouses is identified as a priority - a part of data consolidation that we will discuss later.
Mapping the cloud ecosystem is a key element of our AI strategy because the efficacy of any AI workload is directly related to the quality of the data upon which the workload’s model is trained or augmented. Think back to our earlier foundational discussion of how generative AI acts on enterprise data. Mapping, evolving, and maintaining the organization’s cloud ecosystem map provides the essential high-level technical architecture underpinning our AI strategy.
An ecosystem map is part of a broader approach to what we call ecosystem-oriented architecture (EOA) that we will discuss more deeply as part of the Ecosystem Architecture pillar in the next section.
Programmatic Rigor
Roadmaps don’t get driven on their own, and architecture doesn’t build itself. With an executive vision, actionable roadmap, and ecosystem map in hand, it’s crucial that the organization institute the programmatic rigor required to navigate that roadmap and bring its vision to fruition.
We could have called this “programmatic discipline”. In this dimension, leaders ought to ask themselves if their organization broadly, and their IT teams specifically, are sufficiently focused, possess the rigor and discipline, and operate at a cadence necessary to achieve the milestones they’ve set forth in the time planned. This is program management 101 stuff, so we’ll not rehash it here. Suffice it to say that the organization must act rigorously and consistently to marshal the resources, direct action, monitor progress, and feed lessons learned back into its AI strategy. Organizations that lack this discipline will entirely fail to achieve their vision for AI (or any technology).
As the organization’s AI strategy and program mature it is important to ensure clear lines of communication and feedback are established across organizational lines and stakeholders, ensuring that there is visibility of the program advancement and that success metrics and milestones are being achieved. Regular self-checks are essential to maintain a coherent DevOps strategy. Regular peer reviews, relevant testing, and comprehensive documentation must be standard practices. On the delivery side, IT teams need the discipline to adhere to delivery methodologies, avoiding the temptation to overproduce or create unnecessary deliverables out of an overzealous desire to future-proof. Being future-ready does not mean deploying every conceivable capability just in case it might be needed. Instead, it's about maintaining high standards in delivery and resisting the use of patchwork or temporary solutions that could undermine the very foundation of your platform. Effective programmatic rigor requires not just planning but an ongoing commitment to quality and strategic alignment throughout the development lifecycle.
Center for Enablement
The final dimension in our Strategy and Vision pillar is organizational, putting in place the team or organizational unit required to drive our AI strategy forward. The Center for Enablement (CFE or C4E) concept is rather a departure from IT organizational concepts of old, though, representing a shift from controlling processes to enabling people.
Contrast it to the well-worn “center of excellence” that has historically focused on maintaining standards and enforcing compliance within various technology domains. A Center for Enablement, however, is dynamic: continuously evolving to adapt to new technologies and business needs. It moves beyond reactive governance to more proactively drive the organization’s executive vision for AI.
A well-rounded C4E will broadly focus on the following activities:
• Strategic refresh ensuring that the organization’s AI strategy is continually re-evaluated and refreshed to reflect changes in the technology, business environment, and the performance of the organization’s portfolio of AI initiatives;
• Programmatic rigor, with the C4E taking responsibility for managing the portfolio of AI initiatives across the organization. This must include primary accountability for the achievement of milestones on the actionable roadmap, the development of key workloads, and the organization’s ongoing maturation across the entire AI Strategy Framework;
• Facilitate human connection, creating opportunities for colleagues to collaborate, innovate, and build communities of practice across the organization. This promotes collaboration, networking, brainstorming and building new skills i n accordance with an ever-changing environment. The C4E must enable colleagues to succeed in their use of AI;
• Drive a culture of continuous improvement and innovation, encouraging a mindset of experimentation and learning, where insights are not just consumed but acted upon, iterated, and improved;
• Monitoring and metrics of the organization’s AI initiatives through advanced analytics and AI monitoring itself, identifying patterns and trends in a continuous loop to inform strategy. This must include continual assessment of the organization’s aggregate AI maturity using the AI Maturity Model (discussed in a later section). Learn more about this in the Monitoring and Metrics dimension in the Scaling AI pillar.
The Ecosystem Design Authority (EDA) offers a sound model through which the C4E can facilitate the success of the organization’s AI strategy across technical domains that it may not directly control. Think of the EDA as a collaborative, standing working group; “air traffic control” for the cloud, landing the workloads and technical services of different technology domains in the cloud ecosystem. The Ecosystem Design Authority (EDA) ensures that (a) architecture and technical decisions are aligned with the cloud and AI strategy, and (b) that workloads and technical services are assembled for the benefit of the whole ecosystem, not in service to a specific technical domain.
Figure 8: A notional Ecosystem Design Authority model, though note that the specific choice of technical domains should be tailored for the organization specifically.
A few notes on the EDA model depicted to the right:
“Domains” segment technical disciplines within the ecosystem, and are fluid over time;
Each domain is represented by one cloud solution architect or technical leader, regardless of the number of projects or work streams in the domain;
Domains work together across ecosystem neighborhoods and the ecosystem at large;
The executive leader is a CIO-level or direct report able to make decisions on behalf of the organization;
The following are key roles within the Center for Enablement:
Cloud Strategist is “air traffic control” for the ecosystem, ensuring that technical services fit together and are aligned to strategic priorities;
Enterprise Architect oversees architecture and technical work on a day-to-day basis;
Program Manager is responsible for the non-technical programmatic rigor and execution of the AI strategy.
Here’s where a Center for Enablement becomes truly essential not just to the AI strategy but to the organization’s ecosystem more broadly, where the focus is building adaptable scalable cloud ecosystems that grow with the organization itself.
Navigate to your next chapter…
