The Data Strategy Choice Cascade (2024)

The Data Strategy Choice Cascade (1)

Demystifying Data Strategy

This is part one of a series of articles, in which we will demystify data strategy — an essential component for any organization striving to become data-driven in order to stay competitive in today’s digital world. Use the cheat sheet provided in this article to kick-start your own data strategy development!

  • When organizations strive to become data-driven, they need to transform, which requires a solid data strategy design
  • The renowned Playing to Win strategy framework provides a process to develop, and a format to document and communicate any kind of strategy
  • This article shows how the strategy choice cascade for corporate strategy design can be adopted to develop a data strategy
  • The cheat sheet provided allows data strategy design teams to rapidly apply a practice-proven framework for their own data strategy development thus helping organizations to become data-driven

·
1.1 The Data-Driven Company
1.2 Where is the Challenge?
· 2. Is Data Not Already Handled Within the IT Strategy?
· 3. What Strategy Framework to Use for Designing a Data Strategy?
3.1 The Playing to Win Framework
3.2 Tool 1: The Strategy Process Map
3.3 Tool 2: The Strategy Choice Cascade
3.4 The Playing to Win Framework for Data Strategy Design
·
· 5. The Data Strategy Choice Cascade

5.2 The Data Strategy Choice Cascade
5.3 The Cheat Sheet for the Data Strategy Choice Cascade
5.4 When to Use the Data Strategy Choice Cascade
· 6. Two Example Data Strategies
·
· References

1.1 The Data-Driven Company

Companies big or small, and independent of which industry they are playing in, strive to become data-driven. Regardless of what technology you use or how you name it — Data Science, Business Intelligence (BI), Advanced Analytics, Big Data, Machine Learning or (Generative) Artificial Intelligence (AI) — being data-driven is about leveraging data as a strategic company asset.

The big promises of being data-driven are:

  1. Humans make better decisions: Managers and employees on all organizational levels and across the entire value chain are enabled to consistently make better decisions by combining data & analyses with human common sense.
  2. Computers make automated decisions: Frequent and repeated decisions in operations can be automated.

This, in turn, leads to a long-term competitive edge through:

  • More revenue
  • Less costs, more efficiency and margin
  • Better risk management
  • More innovation and the possibilities of novel business models based on digital products and services

These benefits are realized through individual data uses cases. Some concrete examples are:

  • Targeted customer retention for B2C business: Customer data can be enriched with demographic or market data and is then used to build statistical models (you may call it Machine Learning or AI if you like), to learn rules, which can predict the probability of customer churn. This means, that for each single customer a probability between 0% and 100% is computed, indicating if a customer is likely to terminate his/her contract next month. In combination with a customer (lifetime) value, which also can be deduced from data, the customer service center can leverage this information to make informed decisions about which precise customers to phone up, in order to prevent losing “good” customers. For companies with a large customer base and limited service center resources, this often leads to a significant reduction in churn and hence a long-term measurable increase in revenue.
  • End-to-end product margin analysis: Financial and product data, together with clear cost-allocation rules, can be used to develop a BI dashboard to analyze the margin of individual products. Dashboard users can surf the data and drill-down in various dimensions, such as country, point-of-sales, time or customer type. This allows business users to make informed decisions about which products might be removed from the existing portfolio and hence lead to higher margins.
  • Predictive maintenance to optimize production: Sensor data such as temperature or vibration can be used to predict the probability of machine failure for the near future. Equipped with such insights, maintenance teams can proactively conduct measures to prevent the machine from failing. This reduces repair costs as well as production downtime, which in turn can increase revenue.

The benefits of becoming data-driven appear to be self-evident and there seems to be a general consent that for most companies there is no way around it. A quote translated from [1] nails it:

“Consistent value creation with data is already a decisive competitive advantage for companies today, and in the near future it will even be essential for survival.” [1]

1.2 Where is the Challenge?

If the general benefits of being data-driven are self-evident, where is the catch? Why is not every organization data-driven?

In fact, although the topic is not new — BI, Data Science and Machine Learning have been around for several decades — many companies still struggle to leverage their data and are far away of being data-driven. How is this possible?

To be clear from the start: From my perspective, the actual challenge was and still is rarely a matter of technology, even if so many technology providers like to claim exactly that. Sure, technology is the basis for making data-based decisions – as it is the basis for nearly every business-related activity nowadays. IT systems, tools and algorithms are readily available to execute data-based decision making. Equipping an organization with the right tools is a complicated, but not a complex problem [16], which can be adequately solved with the right knowledge or support. So, what does prevent organizations to continuously innovate, test and implement data use cases — like those above — to leverage data as a strategic asset?

In order to become data-driven, the way in which employees recognize, treat and use data within the daily business needs to radically change.

Being data-driven means, that people in every department must be able to translate critical business problems into analytical questions, which can then be addressed with data and analyses. There needs to be a (data) culture [3] within the organization, that ensures that employees want, can and must use data to improve their daily work. However, changing the behavior of a larger number of people is never a trivial task.

“Organizations need to transform to become data-driven.” [2]

So the real challenge is to sustainably change the way decisions are made in an organization. This cannot be accomplished, by designing and conducting a project, this requires a transformation.

Transforming an organization to become data-driven is a complex – not complicated – challenge for companies. This requires a solid strategic foundation — a data strategy.

However, many organizations do not posses a data strategy, struggle to create one or fear it is a lengthy process. With this article, we want to address this, by demystifying data strategy design and equipping the readers with the adequate tools to make their data strategies a success.

Most often organizations have a corporate strategy, an IT strategy or even a digital strategy. So, where is the need for a dedicated data strategy?

If your organization possesses any strategy, which addresses data as an asset with sufficient depth, you are fine. Sufficient depth means, that fundamental questions regarding data value creation are elaborated and answered, allowing the organization to take concrete actions to leverage data as an asset. However, in my experience, this is rarely the case, and there is often a lack of joint understanding in organizations, how the terms data, digital and technology differ and relate to each other.

Whilst digital strategies usually focus on process digitalization or the design of digital solutions for either internal staff or external customers, IT strategies often focus on system landscape, applications and network infrastructure. Both of them are closely related to generating and utilizing data, but neither digital nor IT strategy is — in its core — concerned with enabling an organization to become data-driven.

The info-graphic provides a more detailed comparison of the terms data, digital and technology:

The Data Strategy Choice Cascade (4)

Consequently, for each organization, it is important to first assess what data related topics are already covered by existing strategies, initiatives and corresponding leaders. Then, if existing strategies do not give a detailed answer to how the organization can leverage data as an asset, it is time to create a dedicated data strategy.

Strategy itself is perhaps one of the most misunderstood concepts [14]. There are many definitions and interpretations. As a consequence, there does not seem to be the one single true approach for designing a data strategy. So where to start?

3.1 The Playing to Win Framework

What we propose here is to select a proven and generic framework, which can be used to develop (any kind of) strategy and apply this to data strategy design. For this, we use the “Playing to Win” (P2W) strategy framework [4], which originates from joint work of Alan G. Lafley, who was former CEO of P&G, and Roger Martin, who worked for Monitor Consulting when starting to develop the framework.

The approach became the standard strategy approach at P&G and has successfully been applied in many industries since. In addition, the P2W framework has constantly been complemented and refined by Roger through a series of Strategy Practitioner Insights [15,17].

The benefits of choosing the P2W approach above others are, that it is widely known and applied and comes with an entire ecosystem of processes, templates and trainings that can be leveraged for designing your data strategy.

In the P2W framework, strategy is defined as follows [4]:

“Strategy is an integrated set of choices that uniquely positions a firm in its industry so as to create sustainable advantage and superior value relative to the competition.” [4]

So strategy is all about choice, making these tough decisions that give you a competitive edge, with no bullet-proof certainty that the choices will turn out to be the right ones. Note, that this quite differs from declaring strategy being a kind of plan [9].

The P2W strategy framework boils down to two core tools for strategy design.

3.2 Tool 1: The Strategy Process Map

The Strategy Process Map is a set of steps guiding the strategy design process [5,6]. In essence, it helps to innovate several scenarios or so called possibilities for what your strategy might look like. These possibilities are subsequently evaluated and compared, so that the strategy team can choose the most promising possibility as final strategy.

It makes use of a rigorous evaluation methodology. This helps to create clarity for each possibility about the underlying assumptions that would have to be true, such that a possibility is considered to be a good strategy.

The Strategy Process Map consists of 7 phases and can be visualized as follows:

The Data Strategy Choice Cascade (5)

3.3 Tool 2: The Strategy Choice Cascade

The Strategy Choice Cascade is the second tool that is used to document essential components of the strategy, which serves as output of the strategy work and is a nice way to visually communicate the strategy to stakeholders [7,8]. It consists of five elements and is often visualized similar to a waterfall:

The Data Strategy Choice Cascade (6)

These five elements of the Strategy Choice Cascade are defined as:

  1. Winning Aspiration: The definition of what winning looks like for your organization.
  2. Where to Play: The playing field on which you will, and will not, choose to compete. Typically, this element comprises five dimensions: i) Geography, ii) Customer, iii) Channel, iv) Offer, v) Stages of Production.
  3. How to Win: Your competitive advantage, how you will win with customers sustainably. This boils down to either lower costs or differentiation.
  4. Must-Have Capabilities: Capabilities required to build a competitive advantage.
  5. Enabling Management Systems: Infrastructure (systems, processes, metrics, norms & culture), which is needed in order to effectively execute this strategy.

The actual heart of the strategy is made of the coherent choices made in box two and three: Where to Play & How to Win.

The boxes of the cascade each stand for a topic for which those tough choices, that we mentioned earlier have to be made.

How the cascade works is best illustrated using a real-world example, which is taken from [9] and describes choices for the corporate strategy of Southwest Airlines.

The Data Strategy Choice Cascade (7)

During the strategy design process, the Strategy Choice Cascade is not only filled once, but is repeatedly utilized at several phases of the Strategy Process Map. For example, every strategic possibility, which is created in the innovation phase, is described using the cascade to build a joint understanding within the strategy team. Also, the current strategy of the organization is described as a starting point for the strategy design process, in order to build a common understanding of the status-quo.

3.4 The Playing to Win Framework for Data Strategy Design

The P2W framework is not limited to the design of corporate strategies, but can be utilized to build any kind of strategy, e.g. for company divisions, functions (such as IT, Marketing or Sales) or even for individuals.

Consequently, it can also be utilized to design data strategies. So, what makes picking the P2W framework for designing a data strategy a particularly good choice?

  • Data strategy design teams often jump into details such as what organizational design to choose (central, decentral or hub-and-spoke), or what technology to apply (data mesh, fabric, lakehouse). These are both part of the last box (Enabling Management Systems) and certainly need to be addressed, but not as the first step. The initial focus should lie on the heart of the data strategy, the where to play and how to win. The P2W framework helps to step back and to focus on the strategic questions to be answered.
  • As the P2W framework is widely spread, there are good chances, that your organization already applies it. If this is the case, it is straightforward to integrate the data strategy with other existing and connected strategies (see also Section 4) and it helps to communicate the data strategy to relevant stakeholders, as the they are familiar with the methodology.
  • Some data strategy approaches seem to me like picking choices from best practices or following check lists to tackle all elements required, lacking real strategic thinking. The P2W framework guides and forces the strategy design team to apply strategic thinking with rigor and creativity.

Whilst the idea of applying the P2W framework for data strategy design is neither rocket science nor new, existing literature [e.g. 12,13] does — to the knowledge of the author — not provide a detailed description of how to adapt the framework for the application to data & AI strategy design in a straightforward manner.

In the remainder of this article, we therefore adopt the Strategy Choice Cascade to data strategy design, equipping data strategy design teams with the tools required to readily apply the P2W framework for their strategy work. This results in a clear definition of what a data strategy is, what it comprises and how it can be documented and communicated.

Applying Data Science, AI or any other data-related technology in a company is not an end in itself. Organizations do well not to start a potentially painful corporate transformation program to become data-driven without a clear business need (although I saw this happening more often than one would think). Hence, there must be a good reason why becoming data-driven is a good idea or even essential for the company to survive, before jumping into designing a data strategy. This is often described as a link between data and business strategy.

The great news is, when using the P2W framework for data strategy design, there are natural ways to provide such a link between data and business, by declaring data management, analytics or AI as must-have capabilities within the corporate strategy, enabling the organization to win.

One example for this can be found in [7], where the Strategy Choice Cascade is formulated for the corporate strategy of OŪRA. This health technology company produces a ring, which captures body data similar to a smart watch. The capabilities are, amongst others, defined as follows:

The Data Strategy Choice Cascade (8)

From this, the need for state-of-the-art data management and data analytics capabilities are apparent.

As the benefits for being data-driven exist in nearly every industry, data management, analytics or AI are nowadays not only must-have capabilities for high-tech companies, such as in the example above. Industries such as manufacturing, finance, energy & utilities, chemicals, logistics, retail and consumer goods increasingly leverage data & analytics to remain competitive.

In whatever industry your company plays, it ideally has already established the need for being data-driven and expressed it as part of the corporate strategy or any other strategy within the organization (e.g. sales or marketing), before you start with your data strategy work.

In many situations, however, such an explicit link between business and data is not given. In such cases, it is important to spend enough time, to identify the underlying business problem, which should be addressed by the data strategy. This is typically done in phase 1 of the Strategy Process Map (cf. Figure 2).

One way to identify the strategic target problem for the data strategy is peeling the onion by conducting a series of interviews with relevant business and management stakeholders. This allows to identify pains & gains — from a business perspective as well as with regards to the management and usage of data in the organization. Once this is done, it is then important to create a consent amongst stakeholders about the problem to be addressed. If there is consent that data is mission critical for the organization, the business need for a data strategy is evident and the corresponding corporate strategy should be updated accordingly.

Once we have established the need for the company to treat data as a valuable asset which needs to be taken care of and which needs to be leveraged, it is time to design a data strategy.

5.1 Translating the Choice Cascade to Data & AI

In order to apply the P2W framework to data strategy design, we first must translate the wording from the corporate context to the data context. So, we start by defining some core elements and rephrase them, where it seems helpful:

  • Offering → Data Offering: This comprises data products, analytics products or services related to data value creation.
  • Customers → Data Customers: People, who are served with the data offering. These are most likely to be internal stakeholders or groups within the company but can also be external users.
  • Company → Data service provider: People, who create the data offering
  • Competition: Alternatives, that data customers might choose, other than the data service provider. This can range from data customers not served at all, serving themselves or using external services.
  • Geography → Focus Areas: Which departments, teams, domains or areas of the organization your data strategy will focus on.
  • Channels → Data Delivery Channels: How data customers get access to the data offering.
  • Stages of production → Data Lifecycle Management: Which stages of the data lifecycle are done in-house or outsourced.

In addition, the strategy definition of the P2W framework can directly be augmented to data strategy:

Data Strategy is an integrated set of choices that uniquely positions a firm in its industry so as to create sustainable advantage and superior value relative to the competition, by leveraging data as an asset.

5.2 The Data Strategy Choice Cascade

Once the core elements have been translated and we have defined what data strategy means to us, we can finally define the Data Strategy Choice Cascade:

  1. Winning Aspiration: The definition of what winning with data, analytics and AI looks like for your organization.
  2. Where to Play: The playing field on which you will, and will not, choose to compete with data, analytics & AI. This typically includes five dimensions: i) Focus Areas, ii) Data Customers, iii) Data Delivery Channels, iv) Data Offerings, v) Data Lifecycle Management.
  3. How to Win: The strategic advantage your data strategy will create. How you will win with data customers sustainably. This boils down to either lower costs or differentiation.
  4. Must-Have Capabilities: The critical data-related capabilities needed to build a competitive advantage.
  5. Enabling Management Systems: Infrastructure (systems, processes, governance, metrics & culture), which is needed in order to support and sustain your data strategy.
The Data Strategy Choice Cascade (9)

Some more detailed thoughts on data monetization options as part of the Where to Play for data strategies can be found in [11].

5.3 The Cheat Sheet for the Data Strategy Choice Cascade

If data strategy is about making choices, what are these choices for each element of the Data Strategy Choice Cascade in detail? The following Cheat Sheet for the Data Strategy Choice Cascade compares and contrasts the cascade for corporate and data strategy in more detail and provides example choices to be made for the data strategy design.

The Data Strategy Choice Cascade (10)

5.4 When to Use the Data Strategy Choice Cascade

Similar to the cascade for corporate strategy, the data strategy choice cascade can be applied in several situations, e.g. it can be utilized to:

  1. Document your current data strategy: This helps to build a joint understanding of the status-quo amongst stakeholders. This is typically done when starting or re-starting the data strategy design process as in phase 1 (identify the problem) of the Strategy Process Map.
  2. Detailing different possibilities: as part of your data strategy design process (phase 3: generate possibilities of the Strategy Process Map). Here it usually suffices to detail the heart of the strategy, the Where to Play and the How to Win.
  3. Documenting your final strategy: After different possibilities for the data strategy have been rigorously assessed, the final strategy can be documented and communicated using the choice cascade. In practice, this documentation is often complemented with more detailed texts and other forms of communication, depending on the particular needs of the audience.
  4. Describing the data strategy of your competitors: The data strategy choice cascade can be used to sketch a rough version of the data strategies of your competitors, which might always be a good idea during the design process of your own data strategy.

In order to illustrate the application of Data Strategy Choice Cascade, we create two (fictitious) possibilities for the data strategy of the company OŪRA. Recall that OŪRA is a health technology company producing a ring, which captures body data similar to a smart watch. A corporate strategy formulation for this company using the P2W framework can be found in [7].

We consider two possibilities, which are quite different, in order to illustrate how the Data Strategy Choice Cascade can be utilized during the data strategy design process to uniformly document, effectively compare and communicate different strategy possibilities.

The first strategic possibility focuses on data-driven features for end-users as external data customers.

The Data Strategy Choice Cascade (11)

The second possibility focuses on analytics for operational excellence of internal data customers.

The Data Strategy Choice Cascade (12)

Note, that the organizational and technological choices are made in the the fifth element — Enabling Management Systems — that certainly depend on the core strategic choices made in the preceding elements. The choice cascade prevents users from falling into the common trap of jumping into technological solutions, before the actual strategy has been defined.

We have seen that the renowned and practice-proven Playing to Win framework for strategy design can be modified for data strategy development leading to the Data Strategy Choice Cascade. The cheat sheet provided can guide data strategy design teams to rapidly apply the framework for their strategy work. Consequently, when it comes to data strategy design, there is no need to re-invent the wheel, but you can stand on the shoulders of giants to ensure your data strategy becomes a success, allowing your company to win.

Is this all? Are we now set to develop our own data strategy?

The cascade is merely a tool, which is used to express many possibilities, that we create during the data strategy design process. If you just use the cascade to write down one possibility for your data strategy, which you think is a good idea, how do you know that this is the best option for your company to win with data?

To find out, you need to apply the Strategy Process Map for your data strategy design, which I will detail in one of my next articles.

[1] Sebastian Wernicke, Data Inspired (2024), book in German language published by Vahlen

[2] Caroline Carruthers and Peter Jackson, Data-Driven Business Transformation (2019), book published by Wiley

[3] Jens Linden, Datenkultur — Was & Warum? (2024), LinkedIn post

[4] A. G. Lafley and Roger L. Martin, Playing to Win (2013), book published by Harvard Business Review Press

[5] Michael Goitein, The “Playing to Win” Framework, Part II — The Strategy Process Map, blog

[6] IDEO U, Strategic Planning: How to Get Started, blog

[7] Michael Goitein, The “Playing to Win” Framework — Part III — The Strategy Choice Cascade, blog

[8] Roger Martin, Decoding the Strategy Choice Cascade (2023), Medium article

[9] Roger Martin, A Plan Is Not a Strategy (2022), video

[10] Jens Linden, Daten — Digital — Technologie: ist das alles das Gleiche? (2024), LinkedIn post

[11] Jens Linden, Datenmonetarisierung — Das Where to Play für Datenstrategien (2024), LinkedIn post

[12] Christena Antony, Beyond the Buzzwords: Crafting a Data Strategy that Works! (2024), LinkedIn article

[13] Deloitte & Google News Initiative, Digital transformation through data: a guide for news and media companies to drive value with data (2019), website

[14] Marc Sniukas, What the heck is strategy anyway? (2024), LinkedIn post

[15] Roger Martin, (Playing to Win) x 5 (2024), Medium article

[16] intrinsify Content Team, So unterscheiden moderne Manager: Komplex vs. Kompliziert (2023), website

[17] Roger Martin, Playing to Win/ Practitioner Insights (2024), website with list of articles

Unless otherwise noted, all images are by the author.

Side note: ChatGPT-4o was used to complement some elements of the cheat sheet example choices and the choices of the example data strategies to ensure anonymity.

The Data Strategy Choice Cascade (2024)
Top Articles
Latest Posts
Recommended Articles
Article information

Author: Roderick King

Last Updated:

Views: 5832

Rating: 4 / 5 (71 voted)

Reviews: 86% of readers found this page helpful

Author information

Name: Roderick King

Birthday: 1997-10-09

Address: 3782 Madge Knoll, East Dudley, MA 63913

Phone: +2521695290067

Job: Customer Sales Coordinator

Hobby: Gunsmithing, Embroidery, Parkour, Kitesurfing, Rock climbing, Sand art, Beekeeping

Introduction: My name is Roderick King, I am a cute, splendid, excited, perfect, gentle, funny, vivacious person who loves writing and wants to share my knowledge and understanding with you.