As we move into the new year this may be a good time to review your data strategy. 

As you do so, you may want to check that you are avoiding some of these common mistakes that often block the adoption of data strategies:

1. Not linking to business outcomes
2. Chasing the latest trends
3. Ignoring the fundamentals
4. Not actionable
5. Not communicated

Let’s take a look at each mistake in more detail!

1. Not linking to business outcomes

Often, data strategies have been treated as the responsibility of IT.  IT departments typically group Strategy and Architecture into a single function, which may mean that the data strategy defined is more aligned with IT needs than with business needs. In fact, it may simply be an extension of the technology strategy.

In one environment, for example, IT’s focus was on reducing its support workload. The data strategy was (simplistically) to move the enterprise data warehouse into the cloud and implement a new self-service BI platform with the goal of making business users responsible for their own reporting. Business users were not engaged in the definition of the strategy, nor were they really involved in defining the requirement for the new cloud-based EDW.

Unsurprisingly, the project ran into difficulties due to a poorly defined requirement and neither business nor IT needs were met.

One very real challenge is that business goals and objectives can sometimes be poorly understood. A business-focused data strategy can be used as a tool to engage with business users, test assumptions and proposed solutions, and engage key stakeholders before substantial technology decisions are made.

This engagement should be supported by strong change management initiatives to help stakeholders understand how investments in data and data management will enable them to meet their goals and objectives, which should in turn help with the adoption of new ideas and platforms.

2. Chasing the latest trends

Quite often we see data strategies that read like a wish list of the latest technology trends. Over the last few years, these trends have included the Cloud, Machine learning/Artificial Intelligence, Blockchain, and the Internet of Things (IoT), to name a few of the most commonly mentioned.

It is reasonable to assume that some of these trending technologies will have a role to play in your business and should form part of your data roadmap and strategy.

But, first and foremost, your strategy should be business-driven.

What business benefit will blockchain, for example, bring to your business? If you are planning to move to the cloud, what should you move first, and why? A shift to the cloud to, for example, deliver more agile analytics will look quite different from a shift to the cloud focused on cost savings.

Your business may not be ready to embrace the latest technologies, may not have the capacity, or the costs may outweigh the potential benefits. By all means, consider how emerging trends will affect your business, and begin to plan for them.

Your strategy in these cases may be focused on learning and understanding, with the goal to consider implementing in the future in order to support business goals.

3. Ignoring the fundamentals

Another common error is to plan big, ambitious “sexy” new data projects without having the fundamentals in place.  These include understanding the data landscape, data governance to manage it, and data quality to ensure data supports business needs.

For example, the move to self-service BI is becoming increasingly popular, but typically businesses don’t know what data they have, or where to find it.  This means that business users don’t have access to the data they need for their BI requirements, or have multiple versions of the same data in different sources, with little guidance on which is correct or the most recent.  Frequently, the data is also not of the required quality to support reporting that is accurate enough to base decisions on.

So a strategy that includes self-service BI – or any other data-heavy requirement – in an environment where the fundamentals are not in place, will be very difficult to implement successfully.

The challenge is to balance the sexy “flavour of the month” goals with the implementation of fundamental data management capabilities that will provide a sound foundation now and for future initiatives. Making small, sustainable gains in improving data integrity, formalising governance, and reducing data debt will pay off and will increase the probability that the next big thing can be realised more quickly.

4. Not actionable

Yes, your strategy should include some futuristic thinking and learning. But it should also enable the business, which means that some of the plan must include implementation.

We would suggest a road map based on a three-month window, to address urgent business issues, a 3-to-12 month window, for strategic projects, and a 12-month plus window for learning and future requirements.

Many strategies are simply aspirational statements, with no thought as to how to actually implement them.  A good strategy takes into consideration where the business needs to go, why this direction is necessary, the benefits it will bring, and plans for how to get there.

Making a strategy actionable requires setting short, medium, and long-term goals, and initiating the necessary projects to achieve these.  It also requires communicating the high-level strategy and end goals to staff, so that they understand what changes are required, and how best to support them.  Many excellent strategies have been sunk by a “but this is how we’ve always done it” mindset amongst staff that didn’t understand the impetus for change.

Your data strategy should inform your technology strategy by prioritising gaps in the data architecture that will provide ongoing value.

5. Not communicated

Many businesses guard their strategies as a closely-held secret, shared only with a few, and taken out once a year to be dusted off and updated.  Unfortunately, this means that when plans are made to implement the strategy, or projects initiated to bring about changes required to support it, they are not supported by staff, and may even be sabotaged.

Most people don’t like change – the unknown is scary and often threatening, so they will do all they can to maintain the status quo.  Not communicating your strategy to staff and selling them on its benefits, is an excellent way to ensure it is never achieved.

Conversely, if staff understand where the business is headed, and why changes are required, particularly how it will benefit them (whether directly or indirectly), they will often come up with new and innovative ideas that may help implement the strategic goals more easily, quickly or cheaply than originally envisaged.

If anything, communicating the benefits to be derived from a new strategy are almost more important than the strategy itself.  It is important that these benefits are concrete and measurable, rather than wishful thinking. For example:

• By using data to inform product development, we expect an increase of approximately 20% in product sales over the next two years.  This increase in revenue will allow us to open several new outlets in under-serviced markets.

One should also recognise that communication cannot be one size fits all. Differing stakeholders and business areas may have very different data subcultures with their own drivers and views of how data must be used. It is important to understand these drivers and position initiatives to support differing perspectives, for example:

• The Chief Marketing Officer might care that the data strategy will inform a better understanding of cross-channel customer interactions
• The Data Privacy Officer might be more interested in understanding how sensitive data will be protected to ensure PoPIA compliance
• and, the Chief Information Officer might be more interested in understanding how the data strategy will help him to identify and reduce duplicated data, and associated costs and risks.

The same deliverable, a data catalog, may be part of the implementation plan for all three of these capabilities, so how we communicate is critical.

__

Your data strategy should be a living document, not too long, not too vague, and not too forward-thinking. Avoiding these common mistakes will help to ensure that this is the case

Photo by Sebastian Arie Voortman on Pexels