An effective data strategy will allow organisations to harness their most valuable asset and drive transformation across the business, for both the employee and consumer.
However, too often, an organisation’s data strategy will fail. There are a number of reasons for this and within this feature, six data leaders will explain these challenges and how to overcome them.
1. Caroline Carruthers
“First, one of the most common reasons for data strategies failing is that they are not data strategies at all. Lots of organisations come up with strategies for data tech or data management plans, but unless you’re looking at data holistically, from the collection of data through to the management of it and everything in between, then you’re essentially building a one-legged stool,” she says.
The other big reason that data strategies fail “is that they don’t underpin the overall business strategy,” continues Carruthers.
“Saying that you’re going to “do data” just for the sake of it is a sure-fire way of your strategy being a huge waste of time. If your data strategy isn’t fully integrated into the wider organisation’s goals and objectives, you’re just putting together a very hollow house of cards that will eventually fall down!”
Maximising the value of data at SSE Energy Services
2. Rich Pugh
“A data strategy should describe how data will be used as a strategic asset, enabling a shift to a more data-centric business model. If you’re not looking to change, you don’t need a “strategy” as such,” he says.
Pugh goes on to outline four steps organisations to ensure their data strategies succeed:
1. Alignment — a data strategy must align to, and enable, your Business Strategy. If it doesn’t describe how data will help to deliver your Business Objectives, it will have little impact.
2. Defence vs Attack — a data strategy must balance data “defence” (govern, manage, secure, protect) and “attack” (use, leverage, model, share, monetise). A “defence” with no “attack” is simply data governance, while an “attack” with no “defence” is unsustainable.
3. Balance — A data strategy must strike a balance across key pillars: data, technology, culture, delivery and capability. Without considering each of these aspects, a data strategy will fail to encompass the overall change needed to succeed.
4. Pragmatism — When creating a data strategy it is easy to leap ahead and get excited about AI & ML. A data strategy should be “pragmatic” above all things, and talk in practical terms about the iterative steps needed to transform towards a data-driven future.
Big Data LDN: Build a data strategy focused on business value
3. Simon Asplen-Taylor
Simon Asplen-Taylor, CEO and founder at Datatick believes “that organisations have the structure wrong and that data scientists are being called in to solve problems they are not necessarily equipped to solve.”
Like Carruthers, Asplen-Taylor suggests that to succeed, any data strategy must be aligned to the business goals. He goes further and says that this strategy should “be written by the CDO who is a business savvy person. Someone who can align all the capabilities of data to the business: increasing revenues, reducing costs, reducing risk, increasing customer and employee satisfaction.”
On a more practical level, Asplen-Taylor explains that all “data sets need to be accessible in order to generate value. It is thought that most data scientists spend only 20% of their actual time on data analysis and 80% of their time finding, cleaning and reorganising huge amounts of data, which is an inefficient data strategy.
“Data sets need to be built, automated and deployed to an environment where the data scientists can access them. The vast majority of company’s data sources that are useful for generating value are within their existing structured systems, so data scientists should first focus their attention on using this data. As the function matures then they can go after different more elusive data sets … but it’s not the starting point. The data sets also need to have proper data governance and data quality applied to them, which is another big advantage of data infrastructure automation: data scientists don’t need to create something from scratch that isn’t fit for purpose.”
4. Mike Kiersey
He points out that too often, “poorly designed data processes will prevent business leaders from reaching their long term goals. A common theme emerging is that not everyone is equally prepared — and a lack of preparedness results in a lack of visibility into critical systems. Businesses are going to have to work quickly to be strategic, customising their plans to their business objectives and culture, as well as any unforeseen circumstances they might be facing, or risk falling behind competitors at a time when they can ill afford it.”
He continues: “One initial failure might be the technologies used to underpin the data strategy in question. For a start, legacy, outdated technologies are unable to support a growing enterprise, and secondly, every piece of this technology needs to be managed.
“This requires the right resources, teams and professionals to implement change and see it through to completion – often resulting in additional costs, and in unnecessary complications to the data landscape, which will drive data strategies off-course.
“However, one of the fundamental obstacles is identifying the people and process issues. Clearly, the difficulty of cultural change is dramatically underestimated in many leading companies. The lack of organisation alignment and cultural resistance is a leading factor in data strategies failing.”
Commenting on the next trap that companies might fall into, Kiersey explains this is “centred on the data itself, and how it’s discovered, stored and managed. Data attributes, including movement, volume, naming conventions and other properties are very sensitive to change. A critical failure would be a lack of due diligence, not providing the right security and therefore not being able to defend against potential breaches and compliance failures.”
A solution to this, turning a failed data strategy into a successful one, could lie with a master data management hub, which organisation’s can use to tie their digital ecosystem together and get complete visibility across the enterprise from a centralised point.
“Machine learning and artificial intelligence modelling can provide critical information to the c-suite such as lead generation changes, third-party data access, breach occurrences and more, and help design and maintain a data strategy that works for the specific business goals,” he adds.
Building a solid data strategy for your organisation
5. Danny Allan
The cloud offers the “temptation of the elastic workloads, infrastructure resiliency, compliance certifications and physical footprint reductions. And while the amount of data being moved onto both structured and unstructured cloud platforms is booming, missteps are easily made.”
When adopting these cloud services a data strategy failure is common because organisation’s “aren’t necessarily familiar with how to properly leverage hybrid cloud and what kind of options are most appropriate to them,” he continues.
“Adopting the wrong kind of solution can lead to availability issues, or even open an organisation up to data exposure or theft. While the first thought is that they must move entirely to cloud, IT leaders should instead be considering what tools are most effective before acting. Costs, security and data demands can be very different between organisations. They need to be considering disaster recovery and availability as a priority and ensure there’s means to restore critical applications and data as rapidly as possible.
“On-premises solutions can sometimes be better suited in some circumstances than moving everything into the cloud. As both require different expertise and skillsets, in addressing the ‘why’ in planning, IT decision makers can be much better prepared.”
The three considerations of data: standardise data, data strategy and data culture
6. Yasmeen Ahmad
She says: “A data strategy starts with the executive arm having a clear vision, business strategy and goals. After all, the data strategy should only exist to support what the business really cares about, whether that be increasing revenue, driving profits, achieving operational efficiencies or managing risk.
“The c-suite set the business goals and must be the champions and advocates for leveraging data as the foundation to driving those goals in a digital, data-driven business environment. A data strategy that does not anchor to business strategy and goals, and without Board and c-suite buy-in, will fail to deliver meaningful returns.”
In order for a data strategy to deliver, Ahmad advises that the “execution requires a coordinated enterprise approach to collect, consolidate and integrate data across business functions, departments and silos, including both internal and external sources of data capture. Bringing together data and ensuring reliability, consistency and quality is a foundation to any data effort.
“However, many traditional, large enterprises struggle with this solid data foundation. After years of business growth, disruption, mergers and acquisitions, autonomous business lines and globalized operations, many businesses find themselves with hundreds, if not thousands, of data silos. The investment, focus and time needed on building a strong data architecture is underestimated and so driving cross-functional, high impact business goals becomes an impossibility.”
She points out that for organisations who do maneuver inconsistent and siloed data successfully and reach the analytics phase, they often find themselves drowning in the operationalisation of insights.
“Data rarely provides an answer and clear direction. Interpretation of data and marrying this back to business reality requires skill that many organizations struggle with. Data literacy and storytelling does not come naturally to business leaders. Time and time again, we see business executives and leaders who take decisions based on gut feel and intuition even when presented with data to the contrary. In order to truly democratize the use of data across the business, we must address the organizational capabilities, as well as culture.”
Looking forward, Ahmad explains that “as companies step up to the reality of a digital, data driven world, where disruption is commonplace, they have moved to become more agile and nimble. This is most prevalent within IT departments, where huge investments have been made available for the latest technology and innovations to drive data strategy.
“However, with hundreds of millions sunk into failed data initiatives involving new technologies that did not deliver, organizations are continuing to clean up their data and technology architecture. With increasing demand for sophisticated analytics and AI tools, supported by scalable data processing and real-time execution, technology teams are struggling to deliver the flexibility, agility, ease of use and innovation support demanded by the business.”