Data Strategy

Bernard Marr

Published by
Kogan Page
ISBN 9780749479855
RRP £19.99
Reviewed by Paul Ramsay
Score

5 out of 10

Bernard is an accomplished author, lightly gliding through several big data scenarios, telling us what he wishes to say, saying it, and then saying it again. A light, easy read with several real-life illustrations from the usual suspects and several others such as Dickey's Barbecue Pit and GolfTEC. From the off, every business is now a data business that needs to capture and manage the right data, transforming it into information to improve its internal operations. Various technical layers about data generated internally as well as bought in, how it is stored and subsequently how the information it contains provide insights. Differences between data, its information, and the strategic business insights it provides are delineated. Tools such as Hadoop and cloud services illustrate technical support for big data. Bernard presents a positive and delightful portfolio of big data, illustrating the core three Vs and extending to a fourth - Veracity, and subsequently addressing data governance.

Internally, business is essentially about optimising complex multi-dimensionality. Regulation enhances some dimensions. A successful data strategy is about envisioning data within the future context, developing out information requirements whilst minimising data. As Barnard says, collecting excess data can be more expensive than just its collection and storage costs. Regulation such as the EU General Data Protection Regulation (GDPR) that has been on the statute book for two years and becomes active 25 May 2018, can incur heavy financial penalties for collecting excess data. GDPR is included here almost as a postscript. Yet GDPR imposes strategic data requirements country by country, even on driverless vehicles. Article 3(2) of this regulation applies to the processing of personal data of any individual "in the EU", including non-EU tourists. Barnard states it applies just to EU citizens.

The legacy challenge is to address big data and integrate it with or replace existing regular data. Information multi-dimensionality results in a diverse array of data strategies. Many of the techniques, categorisations, variety of data analyses and processing illustrated apply equally well to data of either type, regular as well as big. Analytic services such as Amazon Web Services (AWS) and IBM Watson get passing mention.

Developed out of cryptocurrency (for example, Bitcoin), blockchain gets superficial coverage. Everledger, the seminal blockchain diamond identity project is absent. Being a network of distributed nodes critically shareable only on the blockchain, and being intrinsically immutable could have been made clearer. Also immutable blockchain could strategically conflict with GDPR.

Each big organisation will have at least one legacy data warehouse and will be considering progressing to one or more data lakes. Being advised 'keep the data lake in mind as a potential future option' does not offer any clear advice. An old-style data warehouse is briefly touched on as being 'where data is organized in a hierarchical, logical way that is structured and fixed.' Having a data lake used as a free-for-all as suggested could compromise the organisation's ability to have a consistent, valid view across its portfolio.

To summarise this review, parts of this book are very readable, but two or three of today’s hot topics are sadly not given much depth.

It looks like Barnard has a reasonable grasp of data, especially big data, but I believe he should have covered these other topics to a reasonable level, which he covered more in passing than in any depth.

For that reason, I can only give the book 5 out of 10.

Further information: Kogan Page

April 2018