Are you confusing your Data Strategy with your Digital Strategy?

Digital strategy and data strategy, contrary to popular belief, are not the same thing when it comes to how you plan and implement them in your organisation.

Both are significant undertakings. They are linked, but they should be developed individually. This article examines what is in each, and the typical pitfalls you are likely to encountered in designing and implementing a data strategy.

The accelerating pace of technology advancement, high-speed data communications, cloud-based compute power, big data, smart devices, machine learning, cyber threats and, more recently, the step-change in artificial intelligence provides a never-ending list of new things for Boards and business leaders to consider. Do these things present a risk or an opportunity, a fad, a retro-fit or a complete re-think for the business?

Many organisations struggle to formulate a coherent technology strategy because there are so many moving parts across too many dimensions. It’s a challenge to try to cover them all.

From many years of experience in designing data-centric transformation programmes, I strongly recommend an approach that treats data strategy and digital strategy as separate but related components of your business strategy. They complement each other well: wins from a digital strategy really help to build shorter-term confidence in the longer-term changes that a data strategy will bring about.

Digital Strategy

Broadly speaking, a good digital strategy is about enabling movement of data in and out of your organisation and making that data useful for staff and customers in a way that enables them to be more time-efficient. It considers how data:

-? enters the organisation (on paper, manual transactions, batch transfers, real-time feeds)

-? is securely stored, including the underlying data and systems architecture

-? is made available wherever it is needed (via networks, mobility devices, 3rd party interfaces)

-? in a format most suitable to the task it is being used for (words, graphs, pictures)

-? can be accessed and manipulated on a self-serve basis in productivity tools (Excel, BI, AI)

-? is accessed and how people interact with it (the user experience and user interfaces)

-? is managed through workflow so that things don’t get stuck or lost behind the scenes

-? is treated as part of a broader security-aware and cyber-risk-aware culture

A well-executed digital strategy typically improves efficiency, enabling the same tasks to be carried out faster, often at a lower cost-base:

-? Enabling a field-force to go straight to jobs from home by presenting job sheets on a secure mobile device, rather than first travelling to a depot to pick up a paper job sheet, is an excellent example of a good digital strategy outcome. Same job, more efficient, with less hassle.

-? The work of Service NSW in digitising public interactions with Australian state government services is an exemplar of good digital strategy. Many of the back-end IT systems remain pretty much unchanged, but the front-end experience is presented to the user in a clear, joined up and simple manner regardless of the systems, interface and security spaghetti required to get the right data to the right place. Same services, accessed more efficiently, with less hassle.

In both cases, every interaction also invites user feedback to make the digital experience better.

Data Strategy

A good data strategy, on the other hand, is directly linked to the overall business strategy and decisions the business needs to make, or make differently, to enhance the level of maturity in one or more key operational areas or processes.

It should be aligned to the most important strategic business outcomes for customers, staff and investors. It typically involves bringing about change to roles, people skills, processes and the business operating model as well as introducing new types of data and decision-support tools.

For example:

-? in capital investment for building new assets, a data strategy might enable new types of risk-sharing and contracting mechanisms that reduce the cost and time-scale for delivery

-? in infrastructure asset management, a data strategy might enable a shift from time-based asset renewals to risk-based investment decision-making in how and when assets are maintained

-? in infrastructure operations, a data strategy might enable a shift from daily or weekly standard operating pattern to the dynamic management of supply in response to demand

-? in HR, a data strategy might be the enabler for moving from subjective employee performance management to implementing balanced score-card performance management

The list goes on… every part of every organisation usually has opportunity to enhance how it works through better insight from data. Prioritisation and sequencing of data strategy investments should be guided by the business strategy, linking them to the most important strategic outcomes.

Where do organisations usually go wrong with their data strategy?

I have known organisations to take one of three approaches to their data strategy:

1. ‘Big data’ or Internet-of-Things (IoT)-led: Collect as much data as possible, then use whatever business intelligence, AI and visualisation tools are available to analyse it, looking for insight. This can work well with very big homogenous datasets, like retail-sector customer credit card transaction records. But in the infrastructure world, with millions of varied assets, instrumenting those assets, collecting data and storing it is complex and expensive. Analysis of asset condition data typically reveals what could already have been calculated based on the age of an asset, its location and its utilisation. And it comes at great expense.

Very few data-led strategies have succeeded in unlocking significant value in infrastructure organisations. This type of strategy should be reserved for closely monitoring only the most critical assets, or instrumenting a small number of asset types that engineers want to understand better.

2: Application-led: This is probably the most common approach – it is driven by the need to modernise the I.T. estate and consolidate data from multiple antiquated end-of-life applications. Organisations typically implement an Enterprise Resource Planning (ERP) system, maybe also an Enterprise Asset Management (EAM) system and a handful of specialist applications. ERP and EAM implementations are big projects involving a lot of data migration. The assumption in this approach is that the migrated data will be 'clean' and universally more useful.

Failure at the outset to align asset information structures in the EAM system with the chart of accounts in the ERP finance system typically results in the first of many headaches, impairment to regulatory reporting, but the biggest failure is to assume the shiny new systems bring the inbuilt capability for the type of decision-making the business strategy demands.

Most national infrastructure businesses own significant ‘linear’ assets like power lines or roads that ERP and EAM typically don’t support well in terms of the data detail and analytics required. The return on investment is almost always disappointing, and it can take years to become evident.

3. Business-led: This approach is first and foremost about realising the outcomes the organisation is seeking to achieve. It asks, for each major business activity: what are we trying to do differently? What types of data and decision-support capability do we need to enable that? What data elements do we need, and to what level of currency and accuracy? How do we bring different types of data together? How good is the data we have? and where and how do we source any new data we require?

It is often an implicit intent in approach (2), but starting with an unconstrained view of what needs to be changed is far more likely to deliver the desired result.

In large-scale infrastructure organisations the greatest value is usually unlocked by applying insight from data to rebalance elements of asset-related or project-related risk, cost and performance.

This requires carefully managed policy and process changes, and a good understanding of the data that will be needed to support a revised policy approach. This is especially true in safety-critical operational environments where poor quality data, or poor decision-making, can result in asset failures causing injury to staff or to the public.

To carry out the decision-making required by a revised policy sounds simple, but it frequently involves running analytics across a combination of hierarchical, relational, graph, spatial, linear and open data types – that’s not something that most ERP or EAM systems do that well today.

Bringing all these data types together in a meaningful way needs a dedicated approach with the right underlying data specifications, data lifecycle management, decision support tooling and change management approach, which only a well-executed data strategy will identify and quantify.

The business-led approach is by far the most likely to unlock significant value – but it needs commitment from executive management to invest in data as an asset in its own right, requiring the build of an operational data capability with roles and skills that are best implemented separately from typical I.T. responsibilities.

Conclusion

Digital and data strategies should be treated separately within an overall business strategy.

Digital strategy is best sponsored by an I.T. executive director, and data strategy by the business executive for whom effective use of data can deliver the biggest potential benefit – this is typically realised in customer, maintenance or operations divisions.

There are three common approaches to a data strategy: data-led, application-led and business-led.

The business-led approach is the best way forward for any organisation looking to put in place or renew its data strategy. The business-led approach also provides an effective means of recovery from a well-intended but fruitless attempt at an application or data-led investment.

In conclusion, a stand-alone data strategy is vital for any infrastructure organisation looking to unlock the true value of the data it holds. If you haven’t already considered your data strategy, do it now, before the technology landscape gets even more confusing.