Data Strategy Portfolio Management

Data Governance

Data Strategy

Data Products

Data Platform

Data Infrastructure

Adopt an agile, monitorable and value-driven approach to align data delivery activities with strategic business objectives.

Industries:

Finance & Insurance - Retail & FMCG - Energy & Utility - Transportation - Life Science - Industrial

Solutions:

AI - Data Governance - Data Strategy - Data Products - Data Platform - Data Infrastructure

Overview

The role of IT and Data Management functions within companies is changing: they are increasingly considered as sources of added value and less and less as cost centers.

Consequently, the expectations on the organizational units that manage data are constantly growing: the focus is shifting from containing IT spending to new investments in data-driven services, to obtain competitive advantage more quickly than in the past.

In this scenario, “doing right things” becomes more important than “doing things right”.

Combining decision-making speed with the need to invest in activities that produce something truly differentiating for the company suggests focusing more on experimentation and iterative approaches.

Furthermore, a reactive management of business demand, whereby a technical solution is adopted when a need arises, risks being difficult to sustain in a rapidly evolving business.

Quantica has had the opportunity to observe in the field, working with different clients of different sizes, corporate cultures and industries, the need to evolve the method of planning projects in the data sector, as it has been observed that the greater uncertainty and variability of the operating context , also due to external factors, combined with a traditional Plan – Do approach, tends to lead to consequences that hinder the generation of differentiating value, for example:

Difficulty in introducing innovation and seizing the opportunities given by new technologies, with the result of falling behind competitors
Difficulty in removing the technical and cognitive debt accumulated from Business as Usual activities in a short time, with the result of spending more and more resources on maintenance activities
Analysis phases that are excessively and too early compared to the moment of start of the project activities which, consequently, lose their effectiveness as time passes and the context changes for a long time
Prolonged freezes of funds awaiting decisions at top management level, leading to stalemates in the implementation of projects

In particular, in the context of data initiatives, the points listed above, deriving from too static planning or a lack of a structured Portfolio Management activity, often translate into common problems, such as:

Generation of Data Silos

The rigid Plan-Do model inevitably lengthens the analysis phase, but meanwhile, the technical or functional requirements of the business case to be solved often change. To ensure the required delivery speed, during the project, development teams are forced to implement quick and poorly engineered solutions, cutting down on the design phase. Often, existing integration solutions or data structures are replicated due to the lack of tools and processes for sharing knowledge about available data assets or the inability to wait for the establishment of a shared functionality or a reusable data asset. This creates redundancy and potential inconsistencies in multiple versions of the same data asset, with obvious maintenance and governance issues.

Spread of Shadow IT

In the absence of a strategic vision, continuous evolution, and experimentation in both technological areas and in the research and selection of useful data for certain business cases, some business functions are driven to take initiatives on their own, ending up creating stopgap solutions unknown to the rest of the company, not aligned with shared standards, generating additional operational costs and accustoming part of the staff to work with technologies and processes that later become difficult to integrate.

Infrastructure Sizing or Incorrect Technological Choices

Without an iterative adaptation phase of a solution for successive refinements, there are strong risks of making mistakes in sizing the infrastructure or choosing the technological products on which to base the platform, anticipating the final choice too soon. Often, one is forced to define an upfront sizing of the entire infrastructure, based on a hypothetical steady-state of which there are few evaluation elements during the sizing phase. The result is that, in hindsight, many times the initially hypothesized estimate on data volumes, network traffic, and computational load is incorrect: likely, this estimate, on which the provisioning of the infrastructure or the purchase of a license (or an agreement with a cloud provider) was based, led to a waste of funds in underutilized resources or, conversely, to the loss of discount opportunities applied to certain usage scales. Similarly, the selection of the most suitable tools, if not confirmed by implementation evidence on prototypical workloads, risks not being the most optimal for the expected experience.

Loss of Domain Knowledge

An approach that sets up activities for isolated data projects and is not governed by a continuous thread does not favor the continuous consolidation of functional knowledge of the internal IT staff and external consultants on business logic, as the latter are involved in analysis activities and strengthening knowledge only when the need to solve a specific business problem arises (a situation that, in certain cases, may occur sporadically), instead of being continuously engaged in a progressive evolution path of the services offered to the business.

Friction Between IT Function and Business Functions

The delivery of data projects that are not truly effective in improving the user experience and whose value is not perceived by other business functions contributes to creating internal misunderstandings and loss of trust in the value of data.

The traditional management model based on Plan-Do cycles needs to be revised in favor of a new Envision-Evolve model.

It is necessary to adopt a more strategic approach, setting goals and planning continuous investments to evolve solutions, contain technical debt, and improve the service offered to other business functions, in line with their expectations.

The management of the Data & Metadata Management activity portfolio allows for addressing questions such as:

What is the strategic vision that should guide delivery activities in the data realm?
Into which strategic objectives can this vision be broken down?
Which medium-term programs or initiatives are most effective in achieving the strategic objectives?
How can the success of the identified programs be measured?
Which activities might be useful to carry out within the programs?
What outputs and outcomes are produced as value for the company’s stakeholders?
How can the priority of portfolio activities be evaluated?
How can budget spending be committed to activities while mitigating risks?

The capability of Portfolio Data Management becomes fundamental for investing the available budget in the highest-value initiatives, avoiding waste, and ensuring alignment with the strategic objectives that the company as a whole (and consequently, the Data Management function) has set to compete effectively in the market.

Quantyca supports and guides companies in adopting a Data Portfolio Management framework with the following characteristics:

→ Agile: allows for quick and low-impact decision changes, if necessary, to respond to changes in strategic objectives or company priorities.
→ Incremental: organizes activities into modular and sufficiently small delivery units to contain risk, to be planned within a roadmap.
→ Iterative: based on periodic and frequent decision-making and control processes, allowing for governance of the initiated initiatives.
→ Measurable: enables high-level evaluation of the performance of undertaken programs in a structured and quantifiable manner.
→ Value-Oriented: suggests defining project priorities and budget allocation based on the customer value perceived by stakeholders, relative to the cost of realization.

Challenges

The adoption of a framework like the one described is to all intents and purposes a Change Management process, which requires changing some perspectives, for example:

From a focus on the "plan" to a focus on "planning"

Companies operate in a much more uncertain and unpredictable market environment than in the past: in this scenario of rapid changes, the time factor becomes crucial in making or revising strategic choices. The probability of predicting unforeseen events is significantly low: focusing too much on the analysis phase risks slowing down the process and leading to a stalemate. In organizing activities, it is more important to have a method that allows for adapting the planning on the go rather than focusing on an excessively long phase of defining an overall upfront plan.

From a Project Management approach to a Product Management approach

A deliverable can no longer be considered as a static artifact produced once with a certain cost and by a deadline and possibly modified later if a request arises. On the contrary, a deliverable is a dynamic object that can start as a minimal version (Minimum Viable Product) and continuously evolve as part of a roadmap aimed at increasing its value over time for the company.

From a "cost of solutions" perspective to an "investment planning" perspective

To support growth, the evolution of deliverables, and keeping up with the direction of business change, it is necessary to also review traditional budgeting and funding methods. Evaluating the cost of individual solutions without looking ahead limits the ability to ensure continuous service improvement. It is necessary to align the budget usage with the strategic objectives and the programs that achieve them, allocating the overall available budget proportionally to the importance of the strategic objectives they are associated with and the value perceived by stakeholders. This approach helps ensure that all programs in the portfolio are adequately implemented, albeit at different speeds and with different priorities.

Solutions

Perhaps we need to add a caveat to our quest for courageous executives. They need good judgment as well as courage. [...] Your LVT isn’t a planning document that sits on the shelf behind your desk gathering dust. Instead, it is leadership’s vision of the future – from its trunk to its leaves – that everyone in your organization can point to and say, “We are going that way, and I understand why”.
"EDGE: Value-Driven Digital Transformation" by Jim Highsmith, Linda Luu, David Robinson

To implement the Portfolio Data Management capability, there are three useful tools:

• Lean Value Tree (LVT)

• Value-driven Prioritization Matrix

• Measures of Success (MoS)

The following figure illustrates the structure of a Lean Value Tree.

The Lean Value Tree is a tree structure that formalizes and pervasively shares the action strategy of a company, a corporate function or a specific organizational unit, translating the high-level vision into a portfolio of activities to be carried out.

The number of levels and the terminology of the tree nodes can vary and each company can adopt those most suited to its culture and organizational context. Quantica proposes the use of the terms Vision, Goal, Program and Activity, described below.

Vision

Also known as Value Proposition, describes in an unambiguous, concise, yet comprehensive manner the high-level direction in which the company, function, or organizational unit wishes to move to compete effectively in the market or, in the case of a function or organizational unit, to support the company in doing so. The vision should not be too generic but must convey the differentiating traits of the behavior that is desired.

Goals

They articulate the outcomes that are desired to achieve the Vision. These are relatively stable objectives that look at a medium-to-long-term horizon, although they may change over time. The goals should be sufficiently specific but should not already provide the solution or propose deliverables (outputs); rather, they should clarify the ways in which value (outcome) is to be delivered.

Programs

Also known as Initiatives, explicitly state the hypotheses of solutions that are believed to contribute to achieving the outcomes expressed by the Goals and on which it may be worthwhile to invest the budget. They represent ongoing or medium-term delivery initiatives aimed at generating value incrementally through a series of activities that produce deliverables (outputs).

Activities

Represent the atomic tasks included in the programs. Each of them produces one or more deliverables.

The LVT structure is defined as Mutually Exclusive, Collectively Exhaustive (MECE), because:

• it helps to recognize and differentiate delivery initiatives that effectively contribute to achieving strategic objectives, on which to continue or start investing, from those that are not useful to the cause, on which it is not advisable to prolong or confirm the investment;
• it facilitates the mapping of all activities necessary or somehow useful for implementing the programs and, consequently, working towards achieving all the strategic objectives defined as the articulation of the vision.

The Lean Value Tree facilitates the transition to an investment planning culture, in which it is possible to define the budget allocation for the various strategic objectives and, consequently, for the various delivery programs to be undertaken to achieve them. Not all programs will be allocated the same budget share, but by defining a Lean Value Tree, it is ensured that the necessary budget is reserved to implement all the planned programs and to avoid unmanaged situations where the budget is entirely consumed by a single program or maintenance activities (Business as Usual).

The Lean Value Tree is a living, dynamic object and needs to be periodically reviewed, in an agile manner and aligned with the changing priorities of the company, due to strategic changes decided by top management or dictated by external factors. Not all elements of the Lean Value Tree have the same rate of variability: usually, the Vision changes at a very low frequency (once every 1-2 years), the Goals once a year, the Programs every six months. The Activities that make up the programs require more frequent review, usually monthly.

The periodic review activity of Portfolio Data Management is the accountability of a company Steering committee (or function, or organizational unit). In the data domain, the scope of the Portfolio Management activity is restricted to the IT unit or the specific Data Management unit, depending on the company’s organizational structure.

The Value-driven Prioritization Matrix is the tool that allows you to evaluate the activities that make up the Portfolio by priority, based on the value generated for stakeholders, compared to the cost of implementation.

An example of a priority evaluation matrix is shown in the following figure.

Portfolio Management: matrice

The evaluation of activities by priority influences the order in which they are planned within projects. According to an ideal value-driven approach, the planning order of activities should follow the arrow indicated in the figure, i.e. giving priority to the activities that bring greater value and which involve low implementation costs. The overall set of activities that can be carried out is linked to the total budget that has been reserved for the relevant programs.

The Measures of Success are measurable indicators associated with the individual Goals of the Lean Value Tree and are fundamental for periodically monitoring the effectiveness of ongoing delivery programs with respect to the achievement of the strategic objectives to which they are associated.

To have complete monitoring, it is essential to define a minimum set of measures for each of the defined objectives (Minimum Viable MoS set).

The following figure shows the monitoring graph of an MoS.

MoS can be distinguished into two categories:

Leading MoS

Metrics directly and immediately influenced by the execution of individual activities and the production of the related outputs (deliverables). The number of Data Products released, the number of governance policies defined, or the number of platform capabilities implemented are examples of Leading MoS. Core metrics are important for evaluating the value delivered in the short term and with local applicability to a narrow group of stakeholders (for example, a business function that has requested some services).

Lagging MoS

Metrics influenced only indirectly by the execution of individual activities and variables over a broader time horizon as a result of investments and the value brought in the medium term (outcome). Reducing the number of data incidents, improving the service level on deliverables and increasing the number of users of a Data Product are examples of Lagging MoS. Lagging metrics are important to quantify the value provided to a broader set of stakeholders (taking the concept to the extreme, to the entire company) following the choice to invest in strategic programs such as the reduction of technical debt, change initiatives management or investment in automation.

Identifying the appropriate metrics to effectively monitor strategic objectives is an activity that requires some experience. For monitoring to be effective and high-performance, it is essential that the metrics are:

Clear to interpret (simple to calculate and unambiguous)
Quantifiable (in a pure quantitative way or by reducing a qualitative evaluation to a score)
Not polarized (not influenced by architectural choices or by the control system itself)
Relevant (effectively aligned with the defined objectives).

The monitoring of the MoS should be done periodically, at the same time as the Portfolio Management review activity, generally once a month, by the Steering committee of the company function (in the data sector, by the IT unit or by the Data Management unit).

The complete process

Quantyca, through the Strategy & Business Development unit and the CoE of Organizational & Change Governance, has refined the framework definition based on feedback collected from application contexts and consolidated the experience of supporting clients in gradually adopting the described tools.

The complete adoption process involves the following phases:

1 Strategy Definition

Definition of vision, objectives, and success measures

2 Program Mapping

Definition of programs and association with objectives

3 Activity Portfolio Drafting

Identification of elementary activities for each program

4 Periodic Priority Review

Evaluation of activities by priority using a value-cost matrix

5 Periodic Monitoring

Measurement and interpretation of success metrics, defining any corrective actions for the strategy

Benefits

Adopting a value-driven Portfolio Management approach as described entails three main advantages:

Alignment with Stakeholder Value Expectations

Ensuring delivery is closely aligned with the expectations explicitly stated and shared with stakeholders. This reduces the risk of wasting resources on activities that are not actually useful to the cause.

Fair Budget Allocation

Ensuring greater control in covering all strategic objectives with the available budget. This reduces the risk of excessively focusing resources on only a few programs (often dictated by operational needs) at the expense of others that are equally important in contributing to the value.

Investment planning in perspective

Guarantee of investing in a far-sighted way on initiatives aimed at reproducing value over time and not only on the short-term contingent demands of Business as Usual.

Use Case