Date: 13 May 2022
Strategy should always come first. So why do so many data projects begin without a clear vision of the big picture? Why do so many organisations invest in technology solutions before they have a clear understanding of what they want to do with their data?
A data project shouldn’t be any different to any other organisational investment. And the fact that any technology investment should be rooted in your organisational strategy isn’t a controversial idea. After all, all business investment ought to be driven by a clear strategic vision – so that expenditure is aligned with organisational business plan in a way that will deliver on the stated organisational goals.
So, why then, do so many organisations fail to develop a coherent data strategy before spending money on their data projects?
Perhaps it is fear or a lack of understanding: data science is often presented as some sort of “dark art” rather than the methodological science it is. Or perhaps it is a desire to “get started” quickly – commentators often advocate an agile approach to data projects without recognising that there must first be a strategic vision behind agile methodology.
Whatever the truth, the strategic vision must come first.
Why do you need to define a data strategy first?
Let’s consider the risks to your organisation if strategy doesn’t precede investment.
Information overload is a real problem for many organisations. The sheer weight of data generated by today’s digital world presents significant challenges. First and foremost, there is the cost of data storage which can easily get out of hand if not properly managed. However, storage costs pale into significance when we consider the cost of compliance. There are the practical responsibilities of information management, such as responding to data subject access requests. Added to this, organisations must manage the cyber security risks. Failing to fulfil these requirements or falling foul of compliance rules can be very costly indeed. The best way to mitigate all of these obligations and risks is to reduce the data your organisation holds to an absolute minimum. This requires a clear understanding of what data you need to hold and why – in other words, a coherent data strategy.
Given that breaking down siloes of information is frequently cited as a key organisational objective, it’s really important that your data efforts are well targeted. We’ve seen numerous data projects which have ended up creating just another silo of information – exacerbating, rather than solving, the problem.
Without positioning your efforts within your overall goals and acting with clear strategic intent, you risk making your data problems worse – rather than better.
There can sometimes be a tendency to view data architecture as a tactical exercise, rather than something you should consider upfront. However, there are always some important architecture decisions which will need to be taken early on. For this reason, it’s worth involving data architects in your visioning and strategic planning processes.
This way, you can develop architecture principles at an enterprise level. These can then be filtered down throughout the organisation, across all data projects. As long as these principles are sound – and adhered to – you can avoid many practical problems. When you allow individual projects to go off and do their own thing, that’s when you run into problems. Avoid this by taking a strategic approach early on.
How can you properly prioritise projects when you aren’t clear on the overall objective? What criteria are you measuring those individual projects against? While you might have a real desire to get started and an eagerness to deliver practical results quickly, you’re just wasting time if you get started on the wrong thing.
This is especially true in organisations in which an agile approach is advocated. While we are big fans of agile methodology, we also recognise that it can sometimes be used to justify a scattergun or ill-planned approach to projects. While early proof-of-concepts that demonstrate measurable value and return on investment are important to ensure success, these proof-of-concept projects should be carefully chosen within the broader context of your strategic vision.
If you aren’t strategic in your choice of and approach to projects, then you cannot be certain that your efforts are truly targeted on what the organisation wants to achieve.
If you aren’t strategic in your choice of and approach to projects, then you cannot be certain that your efforts are truly targeted on what the organisation wants to achieve.
Given the number of technology vendors pitching in this space, it is unsurprising that business leaders are often tempted to lead with the technology. However, this would be a mistake.
Technology isn’t the difficult part of a data science project – there are usually several solutions that will meet any need. However, if you invest first and ask questions later, a poorly thought-through technology decision can soon become a burden.
In a strategic approach, you should consider which questions you want answers to first. You have to do this before looking around at how you might answer the questions. Or how to present the answers to the people who need them. Otherwise, you can easily end up with some fantastic new software with sleek features and interfaces but with yet-unknown functional incompatibilities that is doomed to become, over time, yet another siloed system.
The importance of a clear, strategic vision
Strategy must come first:
Tactical decisions about:
will all follow from your early strategic work.
If you haven’t yet done the early work to define your organisational strategy, it’s not too late – and our team is on hand to assist.
If you’d like help with strategic thinking or wish to discuss any of the ideas raised in this article, please get in touch.
You can reach us on: 0208 004 3015
Or email us at: [email protected]
What are the risks to your organisation if strategy doesn’t precede investment
Collecting and keeping data beyond purposeful need
Creating more siloes of information, rather than fewer
Legacy or unscalable architecture that makes future analysis more difficult
Companies can create superior value through innovative pricing thinking that combines discount governance, (un)bundling, value-based segmentation, portfolio optimisation, and fundamentally reimagined pricing approaches.
Building a data-driven culture requires companies to recognise that excellence in data quality is no longer an afterthought, but rather the foundation for success in the digital economy. So, why is data quality important?
Far too many organisations still struggle to understand and implement the real essence of a data transformation programme. Too many lack the commitment to execute their strategy at a deeper level.
Tunde is the Co-founder and Director at Trellisi, partnering and advising PE-backed companies, corporate enterprises and scale-ups to accelerate value creation and commercial performance. His key experience includes:
Zoltan is Lead Advisor for Machine Learning and Data Science at Trellisi, specialising in designing and implementing high-value AI/ML solutions for enterprise organisations. His key experience includes:
Simon brings 20+ years of analytics consulting experience translating complex business problems into high-impact marketing insights and predictive modelling solutions. His key experience includes:
Reddy brings 20+ years of Business Intelligence, Analytics and Application Development expertise, specialising in designing and implementing robust data solutions for enterprise organisations. His key experience includes:
John is Lead Advisor for Machine Learning and AI Solutions at Trellisi, specialising in data model design and advanced AI/ML applications across enterprise organisations. His key experience includes:
Shaham brings over two decades of strategic pricing expertise, helping enterprises enhance revenue and profitability through innovative pricing strategies and transformation initiatives.
