AI and predictive analytics are opening new possibilities across natural gas operations, from identifying potential imbalance exposure earlier to helping teams make better use of the operational data already moving through the business.
But what those technologies can accomplish depends in large part on what sits underneath them.
If critical information remains scattered across disconnected systems, spreadsheets, or processes that require routine manual intervention, adding more sophisticated technology doesn’t resolve the underlying issues. It may simply carry them forward.
During the first session of SGA’s From Operational Excellence to Market Leadership series, Scott Vogan, SVP of Sales and Marketing at nGenue, outlined a progression from operational friction to intelligent automation. His framework starts well before AI enters the picture.
AI is Step Four, Not Step One
The potential applications for automation and predictive analytics in natural gas operations are easy to see. Faster access to information, earlier identification of exceptions, and less time spent on repetitive tasks can all create value. But those capabilities depend on the quality of the operational foundation underneath them.
Vogan described a four-layer framework that begins with a single contract-to-cash record, followed by standardized processes. Intelligent automation and predictive insight build on those foundations rather than replace them.
The framework, read from the foundation up:
- One contract-to-cash spine.
Keep the deal, nomination, actual, and invoice on a single record that the desk and the back office both work in, rather than in a data warehouse reconciling several systems after the fact. - Standardize the process.
Define how work should move through the operation and how business rules, contracts, and tariffs are applied. - Apply intelligent automation.
Handle work by exception, so the team touches only what breaks. - Add predictive insight.
Forecast imbalance, load, and margin before the cycle closes, so people can act sooner.
For a natural gas organization, the sequence matters because each step affects what the next one can reliably accomplish. If contract information, nominations, actuals, or other operational data remain fragmented or inconsistent, predictive tools are being asked to draw conclusions from information the organization itself may not yet fully trust. Predictive insight fails whenever the first two layers are broken. Measurement runs alongside all four: a baseline of operational and financial measures shows where to start and whether each layer is producing the intended result.
What the Sequence Looks Like on a Natural Gas Desk
Imbalance management provides a practical example.
A scheduler may be responsible for monitoring hundreds of positions while considering nominations, actuals, telemetry, pipeline constraints, tolerances, tariffs, and contract terms. Traditionally, much of that work can involve reviewing positions and identifying variances that require attention.
Predictive tools offer another option, but only when the underlying information supports them. With reliable operational data and business rules in place, technology can identify which positions are most likely to breach tolerance before the cycle closes, allowing the scheduler to focus on the exceptions carrying the greatest risk.
That changes the business case for the technology.
The value isn’t simply processing hundreds of positions faster. It gives the scheduler an opportunity to intervene before a tolerance breach becomes a penalty or cash-out, while reducing time spent reviewing positions that don’t require immediate attention.
Seen that way, automation isn’t primarily about replacing a manual task. It’s about improving the timing and quality of the decision the task supports.
Better Data Matters When It Changes the Decision
The same principle extends beyond imbalance management.
Consider profitability. If a natural gas organization can’t determine whether a book or position was profitable until month-end, it eventually gets the answer. But transactions and operational decisions occur throughout the month, limiting the organization’s ability to respond to what the information reveals.
Billing offers another example. Recurring EDI exceptions may require manual intervention, delay processing, and ultimately slow invoicing. Simply reporting how many exceptions occurred doesn’t change that outcome. Understanding where and why they are occurring creates an opportunity to address the process producing them.
That distinction is important when evaluating investments in data, automation, and AI. More information isn’t inherently more valuable, and faster processing isn’t necessarily the objective. The business value comes from making useful information available when someone can still act on it.
For one organization, that may mean reducing imbalance penalties or cash-outs. For another, it could mean shortening a billing cycle, reducing reconciliation, improving margin visibility, or allowing the same team to manage greater transaction volume.
Those are operational improvements, but they are also business outcomes.
When “Knowing How It Works” Becomes Part of the Infrastructure
Technology isn’t the only reason to focus on the operational foundation.
Natural gas companies rely on experienced people who understand how the operation works, recognize exceptions, and know what to do when something doesn’t line up. That knowledge is especially valuable in an industry where operational decisions can depend on understanding the relationships among contracts, pipeline conditions, nominations, actuals, and other moving pieces.
The risk emerges when experienced employees routinely become the bridge between fragmented data, systems, and processes.
If someone needs to know which spreadsheet contains missing information, how two systems should be reconciled, or whom to call when the numbers don’t match, part of the organization’s operating process may reside in that person’s experience rather than in the process itself.
As experienced employees retire or move on, the implications extend beyond efficiency. Knowledge becomes harder to transfer, increasing volume can put additional pressure on the people who remain, and adding headcount may become the default response to work that better processes or systems could help absorb.
Strengthening the operational foundation doesn’t diminish the value of experienced people. It can free up more of their time and judgment to apply where their expertise matters most.
Start With the Outcome, Then Work Backward
Growing interest in AI can make the technology itself an easy starting point for modernization conversations. Vogan’s framework suggests a different approach: start with the operational or business outcome the organization needs to change, then work backward to determine what must be true for technology to help change it.
That could mean examining where penalties or cash-outs occur that earlier information might help prevent, where billing takes longer than expected, where recurring exceptions consume staff capacity, or where profitability becomes visible only after the opportunity to respond has passed.
From there, the technology question becomes more specific. Leaders can determine whether the constraint is unreliable data, an inconsistent process, inadequate visibility, a genuinely manual task that could be automated, or some combination of the four.
That also creates a more useful basis for evaluating potential investments. Rather than beginning with what a new technology can do, the organization can ask whether it addresses a problem worth solving and whether the operational foundation is ready to support it.
Build the Foundation Before You Build on It
The next session in SGA’s From Operational Excellence to Market Leadership series moves deeper into the connection between operational performance and financial results.
On October 22, Operational Excellence & Revenue Protection: Turning Process Into Profit will examine where revenue leakage can occur after a contract is signed, across contract execution, billing, and settlements, along with the controls, data and performance measures that can help natural gas organizations identify those losses earlier.
For natural gas leaders considering where automation, AI or better operational data could create value, the starting point may be less about the technology itself and more about the foundation it will be asked to build upon.
