Contributed By: Alisdair McDougall, Commercial Software Product Manager, Schneider electric Energy & Sustainability Services
We’re living in the age of big data. Paired with the call for businesses to become more transparent in their resource consumption through sustainability reporting and goal setting, companies are realizing the need to have robust data programs to accurately report on processes. But this isn’t a new trend: 97.2% of companies are investing in big data and AI to become more data-driven. However, while companies are embracing analytics, the true struggle comes from deriving actionable insights from all the data collected.
How do you know if your company has covered all its bases in the data strategy you’ve set up? Is it time to revisit your strategy to ensure you are getting the most from your data? Below are four tricks to pressure test your strategy and ensure your company is getting the full story from captured data.
1. Collect the right data and define the right KPIs and calculations for your program
Like many things, not all data is created equal. When it comes to data collection, some organizations try to capture every data point under the sun, but many find themselves sinking in the data. Ensure the data that you capture drives insights that support your organization’s strategy and help achieve goals.
Don’t be tempted to create too many key performance indicators (KPIs). Remember: these should be key performance indicators. When defining a KPI, test it by noting some hypotheses and thinking about what it could mean if a KPI changed based off these hypotheses. What investigations would you do and what conclusions could be drawn? This ensures that you will be able to make sense of the data collected and helps identify areas of opportunity and improvement.
Document the processes of your data program, especially when it comes to KPIs or calculated data streams. Strategies change and people come and go. It is important for maintaining continuity in the long run that anyone looking at data or an analysis can quickly and easily understand what they are looking at. Keep on top of data; always ensuring that site, meter and data stream statuses are accurate.
2. Don’t take shortcuts. Capture data available. Automate where possible.
Data collection can be a costly and frustrating exercise. This can lead to companies limiting the amount of data collected to reduce time spent and associated costs. An example of this would be only recording total consumption and cost from an invoice. But there is so much more data behind those numbers that can help paint a clearer picture of your consumption use.
Often, data thought to be the most important will not actually deliver enough insight to drive action. This means people spend time tracing the data point back to its source to get more detailed information to understand the trend. Companies should make sure wherever possible that an invoice is only looked at once, or data captured from a meter or system is recorded once. This will save time and money in the long-run.
To support efficient and accurate data collection, look to automate this process where possible, collecting data direct from source. This will provide cost efficiencies, improve the timeliness of collection and improve accuracy.
Complete data for reporting is also imperative to telling a company’s full data story. Filling gaps in data can be done automatically, for example by using the Gaussian Process Regression (GPR). This process uses machine learning algorithms to predict unknown values in a series. Users can define a level of certainty for the estimation, while providing flexibility of how estimations are made as well as when to fill them. Schneider Electric’s EcoStruxure™ Resource Advisor platform utilizes GPR and machine learning to help companies save countless hours during reporting periods while at the same time allowing for more accurate reporting.
3. Ensure consistency in currency and unit of measurement (UOM) conversion
While it may seem obvious, ensuring appropriate unit of measurement is frequently overlooked. Even experts at NASA once reportedly lost a $125m Mars Orbiter because of a mix-up between imperial and metric units.
Consistency in reporting is required, but more importantly, so is consistency in data collection. Always record data in its local unit of measure and currency. For example, whatever is captured in an invoice should be used in reporting data collection. Make sure your global reporting platform has a robust way of managing UOM and currency conversion to deliver standard reports.
Careful attention must be placed on currency and emission factors, as they change frequently and have a significant impact on reporting. A pro tip is to break out the impact of currency exchange rates or emission factors to make this visible.
4. Spend time on the QA / QC process
A systematic process for data collection and validation significantly improves its quality. However, it will never guarantee 100% accuracy. It is important to put in place a sensible Quality Assurance / Quality Control process in addition. Include individuals close to the data, as they can provide invaluable insight into trends. Use a multi-level review process to ensure checks are appropriate and thorough. Sharing the workload improves results and confidence in the process.
A sure-fire way to ensure you’re getting the most out of your data is by integrating a global data management platform into your strategy. As data transparency becomes more and more mandatory from stakeholders, collecting and reporting accurate data is imperative for all companies. Implementing an enterprise-wide platform enables your company to not only confidently report on its goals, but also provides the opportunity to identify areas of improvement and reduce resource consumption across its footprint. Automated collection and aggregation of data makes what was once invisible, visible to enable better decisions and cost savings.