<span class="text_page_counter">Trang 2</span><div class="page_container" data-page="2">

<i><b>PRAISE FOR DATA AND ANALYTICS STRATEGY</b></i>

<i><b>FOR BUSINESS</b></i>

In my experience, taking people somewhere they’ve not been before requires leadership and trust, whether that’s climbing Everest or making a success of your business through the use of data and

<i>analytics. In Data and Analytics Strategy for Business, Simon </i>

Asplen-Taylor cuts through the jargon and provides a clear route for success.

<b>Kenton Cool, 15 successful Everest summits, one of the world’sgreatest high-altitude mountaineers and leaders, expedition leaderbehind Sir Ranulph Fiennes’s north face of the Eiger ascent andEverest summit</b>

One of the benefits of going digital is that organizations can collect, review and analyse enormous quantities of data. Correctly

interpreting this data provides the intelligence which enables a

business to understand the consumer and marketplace in a completely new way. Successful organizations require a clear data strategy and a disciplined set of operational processes. Simon Asplen-Taylor shows in practical detail how to make this happen in the real world. He

demonstrates that data is key but reveals that an effective data officer never loses sight of the commercial application and human element of the intelligence created.

<b>Kevin Gaskell, serial entrepreneur, and former MD, Porsche GB andBMW GB</b>

Brilliant book! Genuinely the best and most readable book for existing and aspiring CDOs. Every CEO should read the first chapter. Simon Asplen-Taylor has shared his significant expertise to create the go-to

</div><span class="text_page_counter">Trang 3</span><div class="page_container" data-page="3">

<i>data guide for business and data leaders. Data and Analytics Strategy for</i>

<i>Business uses examples from a wide range of organizations to explain</i>

why data can revolutionize a business. A genuinely good read, the book’s structure superbly guides the reader through all aspects of delivering a data and analytics strategy with vital tools and tips. Whether your organization is struggling with trust in its reports or ready to launch the bots, this book is for you.

<b>Nina Monckton, Head of Data, Just Group plc</b>

Businesses operate in an increasingly complex and fast-moving environment, where making the right decision at the right time can mean the difference between winning and losing. Key to this is the successful management and use of data, underpinned by a robust data

<i>strategy. Data and Analytics Strategy for Business provides a structured</i>

approach to show how you can succeed – whether you are just

embarking on your journey, part way through or just fine tuning. The book is full of practical advice, anecdotes and experiences to help you win and not lose.

<b>Carl Bates, former Senior Partner, Data and Analytics, and leader ofthe Ventures practice, Deloitte</b>

One of the ways we can encourage women into data leadership roles is to provide the advice and methods to help remove barriers, while

<i>sponsoring next-generation leaders. In Data and Analytics Strategy for</i>

<i>Business, Simon Asplen-Taylor does just that. He has shared his</i>

experiences and strategies for success to create a level playing field for all data leaders. He also talks specifically about how to build and

leverage the strengths of a diverse team.

<b>Roisin McCarthy, Founder, Women in Data</b>

</div><span class="text_page_counter">Trang 4</span><div class="page_container" data-page="4">

As we strive to gain more value from our data assets, we create more risk, opportunity and motivation for breaches. Simon Asplen-Taylor’s

<i>new book, Data and Analytics Strategy for Business, provides amazing</i>

insight on how you can create more value in your organization’s data while also ensuring its security! Highly recommended.

<b>Ned Finn, CISO, Currys</b>

A very interesting book in such an important and contemporary area of knowledge and skills. Data analytics is not just an area of knowledge that you need to learn more about, but it is considered to be a crucial skill that is required for every business. Therefore, this book is a great addition to the intellectual body of knowledge that can help students, especially those studying post experience executive programmes, such as MBAs and DBAs, to gain clear insights of the key elements of

business analytics and to acquire the required set of skills to compete in the changing world of practice.

<b>Amir Michael, Professor of Accounting and Associate Dean (MBA,DBA), Durham University Business School, UK</b>

The world of a chief data officer (CDO) requires a full understanding of how a business operates, the sector it works within and the people involved. Simon Asplen-Taylor’s book gives a fine insight into the

approaches, decisions and specific actions a CDO can use to bring real value to an organization and make it a critical part of business strategy.

<b>Helen Crooks, Chief Data Officer, Ofgem</b>

Beyond his remarkable expertise across all the complexities of today’s data sphere, this book clearly demonstrates Simon Asplen-Taylor’s mastery of the art of data storytelling. Yet, what makes it even more compelling is that it is so helpfully structured in waves that align to the variety of levels across the entire data maturity spectrum – making it

</div><span class="text_page_counter">Trang 5</span><div class="page_container" data-page="5">

instantly transformative regardless of where an organization currently finds itself in its data journey. This book is an absolute

essential for anyone who wants to successfully leverage the abundant value that can be derived from data and analytics.

<b>Edosa Odaro, Chief Data and Analytics Officer, Tawuniya, and author</b>

<i><b>of Making Data Work</b></i>

</div><span class="text_page_counter">Trang 6</span><div class="page_container" data-page="6">

<b>Data and Analytics Strategy forBusiness</b>

Unlock data assets and increase innovation with a results-driven data strategy

<i>Simon Asplen-Taylor</i>

</div><span class="text_page_counter">Trang 7</span><div class="page_container" data-page="7">

Task: Ten questions to ask about your business Are you worried?

<b>The business case for data</b>

The cost of doing nothing

The value of data is only as good as the value of your business case for it

Identifying the most pressing data problems

</div><span class="text_page_counter">Trang 8</span><div class="page_container" data-page="8">

Defining the CDO role

Task: Setting priorities using the data periodic table Task: Using the data periodic table to design projects Five waves of transformation

Value, build and improve Task: share the story

<b>A team game</b>

Cracking the code as a CDO The résumé problem

Discovering problem-solvers Task: Recruiting for PQ and AQ Beware of ready-made data teams

Anatomy of a quick win

Task: Identifying the right project

Make sure it’s a quick win for everyone Quick wins are not strategic wins

<b>Repeat and learn</b>

Why build a repeat-and-learn culture?Listen to what data is telling you

</div><span class="text_page_counter">Trang 9</span><div class="page_container" data-page="9">

Task: Define a data process

Task: Develop a business change process

Learning and innovating through experimentation References

<b>PART THREE Wave 2: Mature</b>

<b>Data governance</b>

What is data governance?

The importance of accountability

Data stewards, data owners and the data executive Task: Implementing data governance

Further reading

<b>Data quality</b>

The risks of low-quality data The upside of high-quality data The four principles of data quality A data quality strategy

Task: Setting a baseline and a target Task: Build a data quality team

Task: Improving data quality in the short termTask: Improving data quality in the long term

</div><span class="text_page_counter">Trang 10</span><div class="page_container" data-page="10">

Other single views

How do you build an SCV?

Shadow data is the enemy of the SCV Ownership of the SCV

<b>Reports and dashboards</b>

Task: A report audit

From static to dynamic decision support Task: Designing your dashboard

Task: Dashboard implementation From reporting to insight

Task: Information architecture

<b>PART FOUR Wave 3: Industrialize</b>

<b>Automation, automation, automation</b>

What can we automate?

How much can we automate?

Task: The business case for automationTask: Manageable automation projectsWill I use a tool?

</div><span class="text_page_counter">Trang 11</span><div class="page_container" data-page="11">

From quick wins to big wins Be dull and repetitive

Task: Choosing how and when to scale Use your resource multipliers

Task: Implementing your hackathon The dividend from scaling

Best intentions are not optimal Task: Plot a path to optimization Task: Overcoming resistance The limits of automation This is a tipping point References

<b>PART FIVE Wave 4: Realize</b>

<b>The voice of the customer</b>

Hearing their voice

Reasons to use other sources of data Understanding competitors

Customer behaviour Social listening

Task: Applying insights from social listeningTask: Creating a reliable Net Promoter ScoreIntegration

</div><span class="text_page_counter">Trang 12</span><div class="page_container" data-page="12">

Data science, not data magic

<i>How not to do data science</i>

Task: Integrating data science Task: Sustaining data science Embrace the potential for failure References

<b>Sharing data with suppliers and customers</b>

Widespread exposure

Exposing data to business partners and suppliers Blockchain in the supply chain

Sharing improves markets Exposing data to customers Task: Prepare to share

Exposure is inevitable, so do it your way References

<b>PART SIX Wave 5: Differentiate</b>

<b>From data-driven to AI-driven</b>

What does AI do?

A hierarchy of data value The AI journey

Task: Detection

Task: Process automationTask: Improved clustering

</div><span class="text_page_counter">Trang 13</span><div class="page_container" data-page="13">

Data bias

Task: Complex analysis and prediction The limits of AI as a guide or manager

Task: Creating commitment to AI for independent

What is a data product?

A data and analytics centre of excellence (CoE) Task: Creating a data CoE

Three functions of research

A continuous improvement life cycle References

<b>Right leadership, right time</b>

Leading a sustainable data cultureWhich leader are you?

</div><span class="text_page_counter">Trang 14</span><div class="page_container" data-page="14">

<b>LIST OF FIGURES</b>

<b>FIGURE 2.1</b> The investment in data and analytics leverages your other investments

<b>FIGURE 3.1</b> Data periodic table

<b>FIGURE 3.2</b> The five waves to data maturity

<b>FIGURE 3.3</b> Value, Build, Improve (VBI) content of each wave

<b>FIGURE 4.1</b> Data and analytics capability model Value, Build, Improve content of this wave

<b>FIGURE 5.1</b> Quick wins

Value, Build, Improve content of this wave

<b>FIGURE 7.1</b> Data governance maturity model

<b>FIGURE 8.1</b> Data quality dimensions

<b>FIGURE 8.2</b> Customer data quality examples

<b>FIGURE 8.3</b> Data quality maturity model

<b>FIGURE 9.1</b> Overview of single customer view

<b>FIGURE 9.2</b> Single customer view benefits

<b>FIGURE 9.3</b> Shadow data resource

<b>FIGURE 10.1</b> Example dashboard – single customer view

<b>FIGURE 10.2</b> Reports and dashboards maturity model Value, Build, Improve content of this wave

<b>FIGURE 12.1</b> Automation maturity model Value, Build, Improve content of this wave

<b>FIGURE 16.1</b> The data science and AI method Value, Build, Improve content of this wave

<b>FIGURE 18.1</b> Value from data

<b>FIGURE 18.2</b> Data science and AI maturity model

</div><span class="text_page_counter">Trang 15</span><div class="page_container" data-page="15">

<b>LIST OF TABLES</b>

<b>TABLE 3.1</b> Revenue increase

<b>TABLE 3.2</b> Decreased costs

<b>TABLE 3.3</b> Reduced risk

<b>TABLE 3.4</b> Strategy

<b>TABLE 3.5</b> Compliance

<b>TABLE 3.6</b> Enabling capabilities

</div><span class="text_page_counter">Trang 16</span><div class="page_container" data-page="16">

<b>ABOUT THE AUTHOR</b>

Simon Asplen-Taylor is one of the most experienced and

successful data leaders in Europe, having served as chief data officer for several FTSE firms and led some of the largest data-led transformations in Europe.

He specializes in transforming businesses through the use of data, analytics and artificial intelligence whilst delivering

significant upside in revenue, customer satisfaction,

organization efficiency, cost reduction and reduced risk. He has a unique depth and breadth of data experience covering more than 30 years across many industries, having led the data

capabilities at Lloyd’s of London, Tesco, Rackspace, Regus,

BUPA, UBS and Bank of America Merrill Lynch and been a data consulting leader at IBM and Detica.

His major achievements include:

For a UK FTSE 100 financial services organization, delivering a $1 billion per annum contribution to the bottom line and also generating a significant uplift in share price.

For a FTSE 250 business, delivering a 64 per cent

improvement in the company’s margin through the use of data.

Troubleshooting and fixing the data capabilities of two organizations under direct threat from regulators. Being a leader in the data and analytics consulting

practices of BAE systems and IBM – advising clients across

</div><span class="text_page_counter">Trang 17</span><div class="page_container" data-page="17">

financial services, telco, retail, entertainment, media and insurance.

Designing and leading the implementation of the data ecosystem for the UK’s financial regulator, which in turn supported the first prosecution of insider dealing in the UK.

He has an MBA from Durham University, is a Fellow of the British Computer Society and a Fellow of the Royal Statistical Society. He has studied artificial intelligence at MIT.

His awards include being shortlisted for Data Leader of the Year (2019) and DataIQ100 top 100 influential people in data (2020 and 2021), the only fully curated list of influential people in data and analytics. He is a frequent data blogger and regular speaker, panel member and chairman at data events. He is a keen sailor and an avid rugby supporter.

</div><span class="text_page_counter">Trang 18</span><div class="page_container" data-page="18">

<i><b>Why Data and Analytics Strategy for Business?</b></i>

Whether you’re a CEO, CFO, CIO or indeed any C-level role, whether you’re a businessperson who wants to understand more about data strategy so you can decide how to manage it in your organization, or just want oversight then this book will greatly help you not only to understand the challenges of those in your data and analytics teams but also your role in

supporting them. If you’re a chief data and analytics officer, data scientist, data engineer, MBA or business student wanting to understand the growing role of data and analytics in

business, or just data curious, then there is much to be found in this book for you.

This book has been written by a practitioner who’s seen it work and, just as importantly, seen it fail. As the author I only wish this book had been around when I started out, and that’s part of the motivation for writing it.

</div><span class="text_page_counter">Trang 19</span><div class="page_container" data-page="19">

<b>What makes this book different from other databooks?</b>

I’ve thought carefully about the content and at times I feel I’ve given away the ‘crown jewels’. So rest assured I’ve not held back from giving the best advice I can.

Key features include:

Artefacts that will bring significant value to you. I draw your attention to the data periodic table in Chapter 3. It looks simple but it is the culmination of years of

experience and something I use almost daily. Real case studies that bring the strategy to life. Ideas underpinned by solid business theory – with

references to relevant research. If you’re a businessperson you’ll want to know that it’s based on proper research. If you’re a student you’ll want to follow up on the research. Key concepts highlighted throughout the book making it easier to remember and use, and assuming no prior knowledge in the subject.

I’ve followed a maturity model with five waves

throughout the book, to give you a practical approach to follow and to help you measure your progress.

The book is written from the perspective of data being a key business enabler, not a technology capability – the way it should be done.

It’s written by someone who’s made mistakes and is

willing to share the learnings so you don’t make the same ones, and who’s seen all elements of data and analytics –

</div><span class="text_page_counter">Trang 20</span><div class="page_container" data-page="20">

consulted, delivered, owned the capability and been responsible to the board.

I hope you won’t just read this book. I hope you’ll use it!

<b>Follow up</b>

If you have feedback on this book please do share it with me:

If you’re interested in having me consult to your

organization or provide coaching and mentoring services please do get in touch. There is more information

available at www.datatick.co.uk.

</div><span class="text_page_counter">Trang 21</span><div class="page_container" data-page="21">

Along the way I’ve had the privilege to work for some great firms. They’ve had blood, sweat, tears and I hope a lot of value from me. I’ve had some smart colleagues and we’ve had mostly fun times with a few, inevitable, bumpy patches, but we’ve all learned a lot along the way.

I’d specifically like to acknowledge the help and support of: Kogan Page for believing in me, enabling this to become real, and for all the support you’ve provided.

Tim Philips for his support, knowledge and experience, and Catherine Walker for making my illustrations work. Durham University for giving me the best education I could have had, and for giving me the desire to keep learning.

My friends and family – you know who you are.

My three boys William, Alexander and Matthew, who are the main motivation in my life, without whom this would not have happened. I dedicate this book to you.

</div><span class="text_page_counter">Trang 22</span><div class="page_container" data-page="22">

PART ONE

<b>How data and analytics canhelp you grow your business</b>

</div><span class="text_page_counter">Trang 23</span><div class="page_container" data-page="23">

01

</div><span class="text_page_counter">Trang 24</span><div class="page_container" data-page="24">

<b>How can this book help you?</b>

<i>The data your business holds, and acquires through doingoperational and trading activities, is the key to its future. Youmay be surprised to discover how much you can use that data todrive increases in revenue, reduce costs, improve operationalefficiency, reduce risk, and improve customer and employeesatisfaction.</i>

KEY CONCEPTS

<small>Chief data officer (CDO)</small>

<small>Digital transition, digital transformationData and analytics strategy</small>

<small>Single customer view (SCV)Single source of truthShadow data resourceData leakage</small>

‘The world’s most valuable resource is no longer oil, but data’ (Economist, 2017). The article pointed out that Alphabet

(Google’s parent company), Amazon, Apple, Facebook and Microsoft were (and are) the world’s five most valuable listed firms.

</div><span class="text_page_counter">Trang 25</span><div class="page_container" data-page="25">

The assumption that data-equals-oil is broadly true, but not new. It is casually tossed about in meetings or announced with great portent at the beginning of hundreds of conference

Before we discuss the value of data, let’s start by defining what data is for our purposes. You have a business, and that means customers have transacted with it. Every transaction creates data: contracts, orders, invoices, and perhaps refunds and credits. Storing this information means that you have created a database of your customers, containing their names, addresses, mailing preferences, and much more. You have employees too, and so that means another database containing their remuneration, the hours they have worked, where they live and the office they work in, their skills and experience. At some point you purchased goods and services for your

business: raw materials for manufacturing, stationery, energy and water; you spent time on the telephone and rented an office. Records of what you did all become your data. You launched your own products and services that were

researched, developed and released to the market. You wrote press releases, published your accounts, and updated your

websites. That has generated even more data. Your competitors did the same, and you have stored data about that too. Your customers and journalists reviewed your products and

commented on them on social media, which created even more data that you can use.

This, and hundreds more items of information that you

capture every day, is what we mean by data. It matters how we manage and store that data, and also how we use analytics,

</div><span class="text_page_counter">Trang 26</span><div class="page_container" data-page="26">

artificial intelligence (AI), and machine learning to extract the value from it.

In this book I’m not going to make the case for data in itself as an asset – because at this point, I’m assuming that you get it. At this stage, it’s like making the case for railways, or the internet. But despite the fact that every leader in every business

implicitly knows that the company is sitting on a potentially game-changing asset, not many of us really know how to turn that data into value.

Seventy-three per cent of data kept by companies is unused for analytics (Gualtieri, 2016) because most of its potential lies dormant; it is either locked up somewhere it can’t be easily accessed, or it isn’t business-ready.

We will discuss both of those problems at length in this book and outline how to solve them. But think about this for a

second: three-quarters of data is kept, and never used. If data is oil, your chief data officer (CDO) will locate the reserve, estimate its value, and extract and refine it. Not only that, but if you stick to your strategy to create value from data, your CDO will make sure you get better at it. Unlike oil reserves, which are ultimately exhausted, you will never run out of data or, in my experience, ways to use it to help your business. In fact, the volume and variety of data available to a business are growing rapidly. In 2021 it was calculated that the volume of data created, captured, copied and consumed worldwide would be 64 zettabytes (Zb) (Holst, 2021). That’s 64 billion terabytes, or 64 trillion gigabytes – 10 terabytes for every person on the

planet. By 2025, this number is expected to triple.

This book is a guide to how your business, channelled

through a data executive led by the CDO, can leverage its data

</div><span class="text_page_counter">Trang 27</span><div class="page_container" data-page="27">

assets successfully. But, be warned: however tough you think it’s going to be… it’s harder.

There’s a reason why so few businesses do this well. It’s easy to get distracted, misallocate scarce time, resources and

investment, or get stuck in the weeds. This book will also give a structure, five waves of innovation, that can be a template for using data in your business.

CHIEF DATA OFFICER (CDO)

<small>The person responsible for enterprise-wide governance and use of information as anasset. This involves data collection, governance, processing, analysis, business</small>

<small>intelligence, data science and key elements of data-driven AI and other techniques. ACDO usually reports directly to the chief executive officer (CEO).</small>

The chances are that your business is not handling any of its data processes as well as it could. Right now, there’s no shame in that. But it’s costing you lost earnings today, and potentially your future. Your teams today are wasting time finding,

correcting and managing bad data; they have missed

opportunities because they didn’t know they existed; they have made bad decisions because they didn’t know the facts or didn’t trust the data; they are formulating bad strategies or drifting because they have a limited ability to predict what will happen next; they are even risking censure or fines from regulators because of poor data management.

You know your business can do better. This is a great opportunity. It is inspiring and fascinating, but it is also

</div><span class="text_page_counter">Trang 28</span><div class="page_container" data-page="28">

stressful. If data and analytics can make a difference, the window in which to do that is rapidly closing.

<i>The title of this book is Data and Analytics Strategy for</i>

<i>Business because you have one of the greatest challenges that</i>

any business executive can face: the once-in-a-generation shift to a new way of doing business. The data and analytics-driven digital transformation, not least due to the impact of the Covid-19 pandemic, is happening much faster than we expected, and

<i>it requires a strategy to Unlock Data Assets and Increase</i>

DIGITAL TRANSITION, DIGITAL TRANSFORMATION

<small>The process of taking analogue information and translating it to a digital environment.The accompanying transformation involves changing business processes underpinnedby data and analytics to take advantage of the opportunities.</small>

There is no holding pattern. This is both thrilling and extremely stressful. As I will mention many times, this book is not a guide to writing code, or a shopping list for technology. It is intended to be a guide to the projects that your business can launch and complete. Like any executive with a ‘C’ in their job title, a CDO can’t do everything at once or keep everyone happy. This

transformation will need to demonstrate results at every point and change the culture of the business, and that will sometimes cause conflict.

It is daunting. Part of the reason I wrote this book is so that you can learn from my mistakes as a CDO – and hopefully avoid them – while sharing in my successes.

</div><span class="text_page_counter">Trang 29</span><div class="page_container" data-page="29">

If you cannot use your data optimally, firms that have

mastered the ideas in this book and go on to complete the five stages in our journey will challenge your business, and with data to help them, they are likely to win. Data and analytics will help them build better products and services, take your best customers and make them happier, run processes at a lower cost, defend themselves against new competitors, and innovate better, quicker and cheaper, with happier employees.

Your smartest competitors will get on with solving their underlying business problems using data and analytics. They may also be able to invent entirely new ways to do business. Already, the world’s biggest provider of transportation does not own a car. The biggest provider of accommodation does not own a hotel. Supermarkets provide banking services. Our economy is changing, driven by the success of data-driven innovators.

<b>CDOs hold the future of their organizations intheir hands</b>

The strategies to get through this are not secret. There is a community that I have found universally inspiring and supportive who are eager to share their insight: other

experienced CDOs. We’ve been there, and we have the scars to show it. This book would not have been possible without the conversations I have shared with them.

When first-time or would-be CDOs ask me for advice, perhaps the most useful thing I can tell them is not to get lost in the

weeds, and that applies to everyone who gets involved with a data and analytics strategy. Step back, plan, think about how you are going to prioritize, and communicate what you are

</div><span class="text_page_counter">Trang 30</span><div class="page_container" data-page="30">

doing and why. And so before we get into what you could do, how you can do it and what you stand to gain, let’s confront the problems you face head-on.

Below are 10 questions that I ask the board, directors and senior managers when I start a data transformation process for a company. The answer to each is either ‘Yes’, ‘No’, or ‘I don’t know’. Keep track of the answers, and then I’ll explain why I’m asking the questions.

<b>Task: Ten questions to ask about your business</b>

1. Do you invest more than 1 per cent of your revenue in data (excluding the cost of IT infrastructure)?

2. Can you describe a product or project driven specifically by data that had a measurable positive impact on the business?

3. In meetings, do participants use expressions like ‘in my experience…’ or ‘what I feel is…’ as a way to make

4. Do your colleagues argue about whose version of the data, or report, is correct?

5. Have you checked that your customers’ experience of your organization matches your assumptions?

6. Recall the last time there was a crisis in your business. If you had been given an hour to pull some numbers

together for an emergency meeting, would you have been able to find and access the right data?

7. Do you know how much you are paying for data across the whole business?

8. Is there a revenue-generating opportunity that your

colleagues could tell you about that isn’t captured in data?

</div><span class="text_page_counter">Trang 31</span><div class="page_container" data-page="31">

9. Is the collection of data functionally isolated from the process of creating value with it?

10. Can you explain to your customers everything that you do with their data, and would they approve if you did?

<i><b>Do you invest more than 1 per cent of your revenuein data (excluding the cost of IT infrastructure)?</b></i>

The exact amount of your investment doesn’t matter as much as whether you know what you are investing in. I suspect that many managers answer ‘I don’t know’ to this. One obvious reason is that the cost of data is often wrapped into the cost of IT in general. Historically, this made sense: in the early years of the information technology boom you could imagine that

somewhere there was a fixed lump of data, an inventory of every piece of knowledge your organization could assemble, that just had to be interrogated. The constraint was the

investment in systems that could process it.

This is not a useful way to think about the world we now inhabit. On one hand, with the emergence of cloud computing, computing power and storage are services that are as elastic as the need you have for them. Digitization has also created the potential to acquire almost infinite amounts of data to analyse. The data and analytics strategy determines the investment in IT, not the other way round. In fact there is an absolute need to decouple the costs of data and technology.

</div><span class="text_page_counter">Trang 32</span><div class="page_container" data-page="32">

DATA AND ANALYTICS STRATEGY

<small>The choices and priorities that create a course of action to achieve the high-level goalsof the organization. Through the use of all aspects of data – including data generation,data storage, governance, quality, analysis, business intelligence and data science –these goals will create returns or competitive advantage for the business and supportits wider goals.</small>

And, of course, high spend does not mean a high return, although some of the numbers invested by data-driven companies are extraordinary (Krauth, 2018). Google, for example, invests $3.9 billion in data science every year. UPS spends $1 billion. Amazon spends $871 million.

If you don’t know how much you are investing in data, would you know how to find it out? Who would you ask, and do you think you’ll get an informative answer if you do? Often this is a useful exercise. Many businesses have many pockets of data activity that are ineffective because they are buried in silos. Adding up the cost of all these activities may lead you to think: could we be investing this money more effectively?

Which brings us to the next question.

<i><b>Can you describe a product or project driven</b></i>

<i><b>specifically by data that had a measurable positiveimpact on the business?</b></i>

Note the emphasis on measurable results that are attributable to data, rather than technology or improvements in general. If you answered ‘no’, why do you think that is? It might be that

</div><span class="text_page_counter">Trang 33</span><div class="page_container" data-page="33">

there weren’t any projects of this nature. Or it might be that you don’t know about them yet.

We often implicitly know that data will be, or has been, valuable to delivering results – think about a project to cross-sell, or to increase customer satisfaction that increases your share of that customer’s wallet. But often we don’t structure these projects as data-driven, even when they are. We tend to attribute the success to people, products, or an improved process. And, often, that’s part of the story. Being able to tell a story of how data was the catalyst to create value, and being able to quantify that value, has three benefits:

1. It is a way to work out whether you are successfully using data. Can you isolate the role of the data in the business success, and can you express the return on that

2. It is a way to think about how we can innovate. Often it takes a long time, or is not practical, to change a product, process or service. So instead, reframe the question: can we apply better data to existing decision-making, or can we make the data more relevant, to improve a process? 3. Finally, storytelling is a way to share stories of data that

will help others buy into what you want to do. So, if your answer to the question was ‘I don’t know’, this will be the most valuable first step. These do not have to be complex stories – often, the simpler the better, to inspire our

An example: in 1999, a major UK bank acquired an investment and pension provider. In 2010, the data team was given a

strategic goal to use data to increase the size of the business. The bank had 30 million customers, and the investment and

</div><span class="text_page_counter">Trang 34</span><div class="page_container" data-page="34">

pension provider 30,000. And so, by using customer data from bank customers to identify those who would potentially want investment and pension products as well, our team was able to identify and convert 30,000 new customers for the investment and pension provider, doubling revenues accordingly. Data-driven cross-selling doubled the size of that business.

<i><b>In meetings, do participants use expressions like ‘inmy experience…’ or ‘what I feel is…’ as a way to</b></i>

<i><b>make decisions?</b></i>

You’re not alone if they do: Forrester Research reports that only 12 per cent of companies use data-driven intelligence to guide key business functions or corporate strategy (Little, 2016). Listen to two psychologists describe the limitations of what psychologists call heuristic decision-making, but what we know as ‘gut feel’. Gary Klein, a senior scientist at MacroCognition who has analysed the way people make decisions in

high-pressure jobs like firefighting, describes it like this: ‘You need to take your gut feeling as an important data point, but then you have to consciously and deliberately evaluate it, to see if it makes sense.’

Daniel Kahneman, who won the Nobel Prize in Economics for his work on decision-making, trusts his gut even less:

‘Overconfidence is a powerful source of illusions, primarily determined by the quality and coherence of the story that you can construct, not by its validity. If people can construct a

simple and coherent story, they will feel confident regardless of how well grounded it is in reality’ (Kahneman and Klein, 2010).

</div><span class="text_page_counter">Trang 35</span><div class="page_container" data-page="35">

Kahneman concedes there are two contexts in which gut feel has some validity. The first is a predictable situation that is familiar from experience. We all do this every day without noticing: we don’t have time to research every tiny choice we make, so our brain fills in the gaps. Most of the time our instinct makes a good decision for us and if we get it wrong (for

example, picking a film to watch on Netflix and finding it’s rubbish) it doesn’t matter. But this is not the case for most business decisions.

Kahneman agrees with Klein that gut feel is a reasonable starting point in a situation in which the statement is inviting feedback and qualification. What is important is that you test that intuition with data. Think about the opinion as a

hypothesis, one that the CDO can support or dispel by using data. Gut feel is based on past experience, but high-quality data can show that it was based on assumptions that no longer apply (a different market, different customers, different

macroeconomic conditions), or one that adds more detail to what we all assume (showing that the business actually is underperforming in a region, but that this is driven by one office, for example).

Remember, in your data and analytics strategy no decision-making meeting is held without relevant data. You should not trust narratives that overconfident minds have constructed that will lead to bad choices.

Whether you call it heuristics, intuition or gut feel, these

instincts are always with us, and shouldn’t be ignored. They just shouldn’t be used as a way to decide – although our data

limitations mean that’s often what we do. Which brings us to our next question.

</div><span class="text_page_counter">Trang 36</span><div class="page_container" data-page="36">

<i><b>Do your colleagues argue about whose version ofthe data, or report, is correct?</b></i>

In a recent job, I was called in to adjudicate an argument. My CEO had sat in a meeting in which two of his managers fought over some figures for their business unit. The best we can say is that neither of them was trusting his gut: they both brandished reports, with numbers, relating to the same part of the business. But the bottom lines were completely different. He set me the goal of discovering who was telling the truth.

It took four weeks to find the answer that nobody wanted to hear: neither of them. When we dug into it, both had missed out some important costs and made some incorrect assumptions, and so both, in their different ways, had given a too-optimistic picture of the real state of the business, slanted in ways that made them look good. They hadn’t set out to deceive and weren’t aware that that was what they were doing. Neither

report was an effective basis for decision-making but by looking into it we identified some flaws in the business processes and by fixing them, we generated real business value.

This often happens when systems are not integrated, or when there are too many manual processes involved in creating the output. Perhaps the people who create the reports don’t know what the numbers they are adding up really mean. Or people in different parts of the business interpret numbers differently or use the output of different IT systems.

But that doesn’t excuse the practice. Doing this for internal meetings is bad, but having no authoritative source of truth for financial reporting can literally be criminal. The most

destructive example of this must be the bankruptcy of

</div><span class="text_page_counter">Trang 37</span><div class="page_container" data-page="37">

WorldCom. By June 2002, the United States’ second largest long-distance telecommunications company confirmed it had

overstated its earnings, mainly by classifying as capital expenditures those payments it was making for using the communications networks of other companies. It reclassified money it held in reserve as revenues, making it appear it had profits of $1.38 billion. As the SEC reported, the improper accounting entries were easily accomplished because ‘it was apparently considered acceptable for the General Accounting group to make entries of hundreds of millions of dollars with little or no documentation beyond a verbal or an email

directive from senior personnel’. In total, WorldCom made

more than $9 billion in erroneous accounting entries to achieve the impression it was making profits (International Banker, 2021).

At a more mundane level, Gartner Group calculates that poor-quality data drains $15 billion annually from businesses which miss opportunities to sell, or waste cash without realizing it (Moore, 2018).

As a CDO, you might have gone to work at a business that already has good-looking dashboards and elegant

presentations. But does anyone know for sure they can trust the data in them to make decisions, or is it what Sir Alan Sugar, a market trader in London’s Petticoat Lane before he founded Amstrad, used to call the ‘mug’s eyeful’ of technology: good-looking but ultimately irrelevant decoration?

Or, worse, do people create a single truth by averaging several different reports, in the vain hope that they are all equally wrong? Remember, an inaccurate graph will look just as good, and appear just as precise, as an accurate one. Without

</div><span class="text_page_counter">Trang 38</span><div class="page_container" data-page="38">

a single source of truth, you can’t act with confidence on what one piece of data is telling you.

<i><b>Have you checked that your customers’ experienceof your organization matches your assumptions?</b></i>

Pointillist, a firm that specializes in customer experience, surveys businesses every year to discover how well they are doing. In 2020 its survey of 1,050 analytics and customer-care professionals found that 48 per cent thought that lacking a single view of the customer was their number one challenge (Pointillist, 2020).

SINGLE CUSTOMER VIEW (SCV)

<small>The process of collecting data from disparate data sources, then matching and mergingit to form a single, accurate, up-to-date record for each customer. It is also known as a‘Golden ID’ or ‘360-degree customer view’.</small>

A single view of all the transactions and contacts and

communications that you have with a customer is possible if you have a single source of truth about your business, but it is not easy. One of the problems is that what might look like a single, consolidated and complete view of the customer might be missing some vital pieces of data. You cannot discover this by looking only at the data you have. So systematically

surveying your business units can help you find out whether your data represents the whole of the customer’s experience of your business. You might be optimizing the customer

</div><span class="text_page_counter">Trang 39</span><div class="page_container" data-page="39">

experience that your data captures but failing to achieve your wider goals for the business.

SINGLE SOURCE OF TRUTH

<small>Every data element is stored and edited in only one place, with no duplicates. Updatesto the data propagate to the entire system.</small>

This isn’t just about customers: there are two equally important dimensions of business relationships that you can treat the

same way. Do you have a single employee view? A single supplier view? You might think you do, but often you are relying on what your systems tell you, without examining the underlying experiences.

The employment contract is ‘incomplete’, in that it doesn’t tell us much about what makes an employee work hard or be loyal because few of us work in an environment in which our

employer can see everything we do. Therefore, a view of the employee based only on what the contract captures is going to be incomplete. As we move towards a culture in which more employees work from home or wish to work flexibly, being able to structure their contracts and incentives efficiently depends not just on what they can do for their employer but on the full picture of their circumstances, needs, motivations and abilities.

If you have a single view of each supplier, you can identify when that supplier is servicing many parts of your business or where you have multiple suppliers that could be consolidated with a bigger volume discount. Consolidation means you have a huge opportunity to save costs.

</div><span class="text_page_counter">Trang 40</span><div class="page_container" data-page="40">

One of the achievements of a successful data and analytics strategy is a no-surprises culture. No customer, employee or supplier relationship is perfect, but you can aim for an

environment in which none of your colleagues can respond to a problem by saying ‘I had no idea that was a problem’.

<i><b>Recall the last time there was a crisis in your</b></i>

<i><b>business. If you had been given an hour to pull somenumbers together for an emergency meeting, wouldyou have been able to find and access the right</b></i>

We all know by now what it is to nurse a business through a crisis. Everyone whose employer was affected by Covid-19 can recall the scrambling and off-the-cuff decision-making. Every day involved rapid replanning, most often around protecting cashflow. And so suddenly people are asking for different data, because there are different decisions to be made. They want it now, because tomorrow is too late.

So, imagine that you need to analyse one of these ideas, and to do that you need data about the current performance of the business. How much detective work do you need to do to find the numbers you need? Will you need to ask someone else to locate it? How many manual processes will you go through to extract and consolidate those statistics?

Often, it is quite a lot of work to locate this data. This happens because we too often get used to the tedium of assembling and curating data. Each quarter, season or month there are routine or repeated tasks that you and your colleagues complete that require the same types of data, just with different dates or

</div>