View all blog posts
3 months ago
  • Data 101

Six key steps to building a data-driven energy business

Zeno Technologies ⓒ Zeno Technologies. All rights reserved.
  • Twitter
  • Linkedin

Six key steps to building a data-driven energy business

Written By Joel Carusone
  • Data 101

All modern industries rely on data to drive enhanced business performance. Leading businesses in these industries have moved past subjective decision-making and now routinely utilize sophisticated modeling that leverages data from numerous sources to make smarter, more timely decisions.

So how can learnings from these industries be applied to the energy sector, which is struggling to make sense of a torrent of different data to inform better financial decisions? In this blog Joel Carusone, Zeno’s CTO sets out the six key steps that energy businesses should take to become truly data-driven.

 

Step one: Decide on the questions you are trying to answer

Becoming data-driven historically meant building monolithic data warehouses, pulling in any and all available data, attempting to normalize it to a common format, layering on validation logic, and then blending it all while maintaining strict controls over how it’s secured, accessed, and updated. This approach has proven to be extremely time-consuming, and often far too costly and restrictive. Furthermore, despite best intentions, planning, and tools, these projects still experience a high failure rate as many organizations lack dedicated expertise in data management and governance, so long-term system maintenance falls short.

“What we have learned is that businesses only require a small percentage of their data to gain valuable insights. A far better strategy is to extract the only key data necessary for gaining specific predetermined insights, and then incorporate additional data over time as it is needed.”

How do you determine what data to bring in? Start by identifying the questions you are trying to answer. Identify the key sources of information and the key data points within those sources necessary to answer these questions.

Once you have incorporated the data necessary to answer your first set of questions, identify a new set of questions to answer and repeat the process. Each iteration will bring valuable knowledge that will help inform the next set of questions to answer while also preventing data that will sit unused from adding additional overhead and burden.

 

Step two: Verify you have healthy data

It’s impossible to make critical decisions based on poor-quality data. At best, businesses using inaccurate data rightly mistrust what that data is telling them and often revert to subjective decision-making. In the worst case, they believe the inaccurate data and use it to make misinformed decisions, sometimes with serious consequences.

‘Garbage in, garbage out’.

So, how can you ensure that you’re avoiding this common trap, and running your business on high-quality data?

The answer lies in utilizing combinations of data quality rules, ML/AI, and data quarantining to govern long-term data management.

As new data points are added, they are not only evaluated against your pre-defined rules but are also compared to other values in your dataset or 3rd-party data sets, with outliers flagged for more scrutiny. In more advanced systems, data quality is often given a health score to assess and rate its accuracy which should continually improve over time.

 

Step three: Derive intelligence from information

In an industry where there may be a dozen sources for the same piece of information, how do you know which ones to use and when to use them? The answer lies in a new data strategy that uses dynamic blending.

“To get an accurate representation of energy assets, you need to blend data in a dynamic way.”

Unlike traditional ETL approaches whereby data is blended on ingestion and changes to blending rules require all historical data to be reprocessed, the new approach is to create composite records utilizing dynamic rulesets. These records are generated at run time and exist for short periods of time. Dynamic rulesets effectively ‘tell’ the data platform which data source to use under which circumstance. This means different departments can look at data through a lens that makes sense for them, while also having the ability to change that lens without impacting other users over time.

 

Step four: Visualize what the data is telling you

Understanding what numbers mean requires translating data points into visualizations. In the past, that often meant extracting data out of a data platform, loading it into Excel, and then reloading it into visualization tools like Spotfire or Tableau.

“Human beings are not hard-wired to absorb and interpret large volumes of numbers.”

With a data platform specific to the energy industry that includes out-of-the-box visualization templates, users can move quickly from analysis to insights, slicing and dicing data directly within the platform, to gain insights that are most important to their particular business.

 

Step five: Collaborate to build consensus and inform insights

As the energy industry collectively looks to take steps forward, we increasingly recognize the need to work collaboratively and cross-functionally and move away from siloed team structures many of us have experienced in the past. Making critical decisions requires input from multiple internal departments and key external partners, which requires a fundamentally different way of interacting. So, what is the solution?

“The answer is to maintain a master view of critical business data in a single platform and, leveraging advanced controls and permissions, allow multiple teams to easily engage in cross-functional planning and analysis.”

It is important to let cross-functional teams see and work on the same data at the same time, while also allowing the organization to capture changes through time-stamped version controls so they can easily audit adjustments, retain tribal knowledge, and even revert to prior versions if necessary.

 

Step six: Understand and act on your key business drivers

Most data platforms let you collate numbers and visualize data. Good platforms deliver additional value by helping your business understand where its sensitivities lie. What would a $5 change in price mean in regard to the profitability of a specific energy asset? Would a 5% production decrease make a project go upside down?

“The best platforms help decision-makers truly understand and act on their key business drivers.”

Knowing your key business drivers lets you focus on what’s important, and fine-tune them to achieve outsized business results. Simply put, a data-driven approach offers a powerful competitive advantage over industry peers.

 

Summing up

Following these six key steps provides energy businesses with a new, workable approach to become truly data-driven. In so doing, they finally have the ability to make the smarter, faster decisions needed to enhance their business performance and thrive in the new market realities they face.

 

Zeno

Zeno’s Energy Operating System was built from the ground up to connect the entire business through data, surfacing key insights for smarter, faster decision-making. Learn how Zeno helps businesses thrive in the new market realities of the Production Era by getting in touch.

 

 

Authors

Joel Carusone
Joel Carusone
Co-founder and CTO, Zeno Technologies

RELATED POSTS

What ‘Moneyball’ can teach energy businesses about data
7 months ago

What ‘Moneyball’ can teach energy businesses about data

Read More

Ready to get started?