Why Data Is Overrated
Data is widely considered to be at the heart of IT strategy. Buzzwords abound like ‘data driven decisions’, but data has no inherent value – only in the way that it is used.
To understand this more, let us look at a couple of key definitions:
Data – a sequence of one or more symbols, normally stored for computer processing
Information – imparting knowledge
Insight – gaining an accurate and deep understanding of a complicated situation
We can thereby conclude that whilst data is just a series of numbers on a page (or in computer memory), we need to turn it into information to understand what happened and into insight to understand why.
It is only when we understand why that we can take action to maximise or minimise the reoccurrence of the situation this data denotes.
The Journey to Actionable Insights
The journey from Data to Actionable Insight can be thought of in two stages:
- turning data into information
- turning information into insight
Something lost to many is that the process of collecting data matters. For data to provide information when analysed, data must be descriptive, and for it to be descriptive at scale it must be well structured and conform to certain standards. Data without context provides no information, and the goal of structuring that data is that it can be analysed quickly, and efficiently (automated) to provide knowledge.
The goal of insight is to help you change something in a favourable way, and this relies on the predictive power of knowing not only how, but also why things happen. Creating insight (the ‘why’) relies not only on establishing correlation but causation. Once causation is understood then prediction is only one step away.
Key Considerations to Ensure That You Can Draw Insight
Whilst the path from data to insight isn’t always easy, there are a few simple and practical tips to consider:
- Define A Structured Data Model – The Foundation Data Model (FDM) acts as a skeleton around which you structure and build your reporting. Each business dimension (worktag) should be a meaningful way of analysing data, whether that is a region, business unit, or something highly specific to your industry or organisation. Just as important is that they only mean one thing. A dimension that is used to mean multiple things may well end up as both rows and columns in a table, and not only does that make analysis hard, but it makes insight next to impossible. Definitions of what a dimension does and doesn’t mean will help you to ensure that your structure remains intact.
- Use Worktags Consistently – If two transactions are captured with different detail (e.g. one doesn’t contain a Worktag type that the other does) then your data isn’t normalised. If this happens then you are no longer comparing like for like, and the conclusions that you draw may have their integrity compromised. Use custom validations and configure worktag usage such that worktags must be entered consistently and prohibit combinations that don’t make sense.
- Create a Source of Truth – You don’t always need to report directly from your system of record (SoR), but you do need to be confident that it is a Source of Truth (SoT). Data in a SoT will reflect the same attributes and the same conclusions as the SoR. If there are many different SoR’s within your organisation ensuring a consistent SoT involves ensuring full traceability back to the original transaction.
- Utilize Drill-Down Reporting – Good insight relies on thoughtfully designed reporting that prompts curiosity to dig a little deeper. Reports with native drill through to underlying transactions ensure that no summarisation happens inadvertently.
- Get Serious About Governance – Business systems need to evolve and change as the organisation does, so it is inevitable that your needs will change over time. The job of systems governance is to ensure that the structure expands sympathetically and consistently with the designed schema. The purpose of governance is to maintain the integrity of the systems that provides the confidence in your insights.
Putting It All Together
Historically, Finance, HR and Business Operational data used to reside in different systems, bound by different data structures and standards. Analysing anyone in isolation was relatively easy but led to siloed views of the business and didn’t offer much in terms of insight or predicative power.
Now, with the power of the Workday platform, it is easy to get all this data in one place that conforms to the same standard. Whether that is using PRISM Analytics to bring operational data into your platform environment or Adaptive Planning to drive plans and strategic interventions.
In most organisations, it is the ratios of metrics from across operational, financial and HR datasets that best describe the causal link between actions and results. To draw these conclusions, it is critical to be able to align relevant data points from the data sets, and this is made possible by following the above guidance.
This post was written by Collaborative Solutions, they are an exhibitor on the HRTech247 Workday Partners floor in the Partners Hall here.