Data projects often represent a complex piece of work that organisations undertake as they shift to modern and sophisticated solutions to serve their use of data. These projects face a range of challenges which impact both the delivery journey and the outcome for customers. Whether run as a specific data enhancement project or as a data and reporting stream of a larger initiative, data outputs delivered for consumption are the ultimate outcome that justifies the project’s effort.
In a recent installment of our ongoing ‘Lunch & Learn’ series, we discussed some of the challenges of data projects and how to address them to ensure a meaningful outcome for the end customer. Our discussion was framed around the process of delivering successful consumption solutions.
Data projects are not immune to the typical obstacles that are present in many technology projects, including resourcing, access to subject matter experts, missing or incomplete documentation or strong sponsorship.
Taking an end-to-end view, we identified four key challenges on top of the typical ones that shape the success of data projects:
- Level of engagement and approach
- Data acquisition approach
- Understanding the data required for decisions
- Data consumption patterns
The first two challenges focus on core foundational elements of data projects and the last two focus on the actual consumption needs of data. Together, they strongly influence the overall outcome of data projects.
Begin with the end in mind
The first challenge is the early, and direct engagement with data stakeholders at the initiation of the solution life-cycle. It is our experience that this doesn’t always happen appropriately, thus leading to data stream estimates with minimal stakeholder engagement. A data project team that is not well engaged with stakeholders early will find itself in a difficult situation to deliver a quality outcome. Consequently, this can lead to convoluted data solutions and create a disconnect between the delivery team and stakeholders, resulting in a tense delivery environment.
If a delivery team does not have sufficient time or information to undertake proper data estimates, our recommendation is to create adequate flexibility in the following delivery stages for the scope to be assessed fully as the requirements are refined.
Indeed, as data outputs are generally at the end of the project deliverables, it is helpful to recall the words of Dr. Stephen R. Covey; “Begin with the end in mind”. The reasons above indicate that it is vital for data teams to engage with stakeholders early in a project to ensure smoother estimation and provision of optimal data deliverables for the customer.
Explore data acquisition options
The next challenge related to the success of a data project is that of data acquisition. As storage options have expanded with cloud computing and flexible architectures, the opportunities to obtain more data are ever present. As a philosophy and subject to project constraints, we recommend acquiring as much data through as many data sources as possible. Although possibly driving up initial estimates, the benefits of increased data can provide value later in the project. A larger amount of data acquired offers more options for the delivery of data objects. The approach should therefore be flexible based on the scale of the initiative being delivered.
Think above and beyond immediate consumption
The third challenge in our discussion pertains to shaping large amounts of data and incorporating them into deliverables of value for customers. Our recommendation is to approach this through an understanding of the key drivers and the data relationships from a build perspective. This is pivotal as it enables the data objects to be constructed in a way that will support immediate consumption needs, plus enable re-use and easier extension into the future. Tackling this challenge requires seamless collaboration with subject matter experts and consideration on how the data construct and outputs will be used beyond the immediate scope. This ensures that the data build aligns with business needs and contains key elements required to support strategic decision-making.
A business’ ability to consume and extract key information from a deliverable; for instance, a data report, and connect this information to its decisions is a key metric upon which the success of a data project is determined. Hence, thinking beyond immediate data consumption needs and establishing the foundation which supports future use is a success criteria of data projects.
Build to usage patterns and advocate less is more
The final challenge in data projects relates to democratised access to data with greater intent on behalf of users to draw on data directly. Specifically, for data consumption, a project may need to:
- Create mechanisms for users to access, use and download data (per data governance frameworks).
- Enable efficient utilisation of data without the need for “re-packaging” reporting content.
- Provide visualisation/graphical data that enables decision-making natively in the chosen toolset.
In our experience, at the scoping stage of data projects a high number of data consumption outputs may be identified in scope. If demand for structured reports is high, we recommend constructively challenging the number of reports or visualisations to be built. This can be achieved through prioritisation of key outputs and delving further into the planned use cases to identify optimal data access paths. Evidence shows that usage patterns of reports are concentrated on a core group of reports with others attracting low usage. Therefore, applying a “less is more” philosophy will pay dividends and help avoid effort on outputs that deliver low value as structured reports.
Regardless of the reporting tool, the aim should be to build consumption outputs that capitalise on the functionality strengths of the tool and provide required information to users as efficiently as possible. In doing so, the data project must provide users with a means to obtain and consume data easily in a way that aligns with their use cases, including any of the actions or decisions to be made from using the data.
Have you experienced some of these challenges in data projects? Get in touch if you want to discuss in more details by contacting us here.
