Blog

Data Analytics Marathon: Crossing the Finish Line

The story of Pheidippides inspired the origin of the modern marathon.

The Athenian messenger ran about 42 km from Marathon to Athens to deliver the news that the Persians were defeated by the Greeks in the Battle of Marathon in 490 B.C., shouting “Nenikekamen” (“We won!”) and dying shortly after. While the story of Pheidippides demonstrates to this day a remarkable achievement, anyone who has run a marathon can tell you that in order to cross the finish line you need a good strategy, a race plan, adequate training, and lastly a strong execution. Similarly, when it comes to data analytics, organisations have their own “marathon” challenge to complete in order to get to the finish line.

OilX is a pioneering oil data analytics firm, working with major commodity trading organisations and supporting them on their path towards data-driven decision making. OilX’s data science framework combines the latest AI technology with fundamental oil market datasets to empower a “quantamental” style of analytics and trading. The path towards that goal can be characterised as a data analytics marathon.

At OilX we observe that companies start the marathon like a sprint, and then find it difficult to finish it. This leads to frustration on all sides; managers feel their investment in data analytics has not delivered its expected value, while data scientists get frustrated when the role promised in the job description does not materialise. Most of the time, the dissatisfaction can be traced back to a missing focus on the finish.

Data analytics can provide every company with valuable return on investment and competitive advantages when performed correctly. For example, refineries are scrambling to get an edge over traditional ways of sourcing crudes and are starting to tap into digital technologies and alternative datasets to drive decision making. The process can be drawn out and tedious if it is not organised with a clear set of goals. Brent Dykes, the Chief Data Storyteller at AnalyticsHero, describes the challenge as follows: "There is a significant drop off at the last mile in the race where analytics teams perform analysis, share their insights and then implement changes to optimize the business. Most companies have no problems with the start of the data analytics marathon, but many of them are not completing the entire race. In fact, they are more likely to bail and start a new race than see an entire marathon through to the end. As a result, they are constantly running but never finishing anything."

Source: Six stages of a Data Analytics Marathon, by Brent Dykes in Forbes, adapted by OilX.
Breaking Down the Data Analytics Marathon

We consider six steps to be crucial to complete the data analytics marathon, and giving your business the upper hand. The first three - data collection, preparation, and visualisation - are the foundation of the data analytics marathon. These must be performed with diligence and precision, or else the race will never be finished. At this stage, envisioning the end of the race can be difficult, especially when working with large and complex datasets, but it is imperative to stay aligned with core goals. At OilX we apply these steps as follows:

1. Data Collection: OilX’s sources innovative, alternative datasets from around the globe in order to provide real-time, accurate views on oil markets that are all updated daily. Whether we are monitoring global oil production utilising satellite data, tracking traffic or tankers, our data collection methodology needs to be robust, relevant and credible. Collecting and constructing data from the highest quality sources adheres to our commitment to excellence.

2. Data Preparation: Implementing data science into oil trading strategies is one the biggest trends in the market right now, as trading desks shift to a data driven approach. With multidisciplinary teams in Athens, Gdansk, London, and Vienna, OilX’s data scientists work together to transform data from across the globe into a coherent format. Whether you are looking at global oil demand or global refinery throughput, data preparation and curation by data scientists and oil market experts is required to analyse and leverage the data.

3. Data Visualisation: We display interactive and intuitive charts, graphs, and diagrams that provide actionable insight. Whether you need an OPEC supply trendline, global oil demand, or the seasonality of refined products balances, our wide range of models allows for data analysis to be conducted in a variety of ways. That allows us to provide a coherent and robust visualisation of the global oil supply-demand balances, all the way from “well-to-wheel”. Data visualisations such as below are crucial in the data analytics marathon, especially for completing the last mile. Failing to make the data tangible to users can be misleading or lead to missed opportunities.

Source: Crude Oil Supply-Demand Balance from the OilX Platform
The Last Mile

4. Data Analysis: At OilX ensure users are able to see data in context - otherwise, a visualisation, table, or statistic can often be meaningless. OilX provides oil traders and analysts with indispensable market analysis, key trends, and main outliers well ahead of official data releases and relative to consensus estimates, allowing to make decisions faster with more certainty.

5. Actionable Insight: Through the combination of research, market intelligence, and performing the core fundamental oil analytics tasks to the highest standards, we free up time for data-driven decision making and trade idea generation. We also help our users to blend OilX data with their own proprietary datasets to create new, tangible insights on and around their own data to help them navigate markets with confidence.

6. Take action: The last - yet most important - step is to decide which actionable insights should be pursued and prioritised. In some cases, someone might want to deploy a single pilot first to verify the results before making alterations to their process. Alternatively, one can deploy a series of insights and run them as a portfolio of ideas or trading strategies. Ultimately, taking action should be the main goal for every company entering the data analytics marathon.

By adhering to these six steps, your chances of deriving value and finishing the data analytics marathon will increase. From here, projects can be completed with more confidence and at higher success rates. At OilX, we turbocharge the analytical efforts of our clients and help them finish each marathon, so that they can start the next.

Sources

Dykes, Brent. Forbes Magazine (January 13th, 2022) | Data Analytics Marathon: Why Your Organization Must Focus on the Finish.
Frost & Sullivan (August 10th, 2022) | OilX Applauded by Frost & Sullivan for Emerging the World’s First Digital Oil Analyst