At the Saga SÖKKVABEKKR conference in Montréal in 2025, Odin Ai presented results showing how C-Factor can explain production performance, landing depth, and the effect of frac execution on well outcomes. Since then, the same underlying approach has been reproduced in the Haynesville, with new projects in development across U.S. plays including the Wolfcamp, Eagle Ford, Woodford, and Marcellus.
C-Factor did more than forecast. It diagnosed why performance changed.
The Halo analysis identified two major drivers of the 400% difference in production between the SW and NE part of their field: rate and landing depth.
Rate and rate timing played a signficant role in production outcome, accounting for at least 50% of the difference.
It clearly identified that as the well porpoised away from the Montney S2, C-Factor fell, and as it moved toward the S2, C-Factor rose.
That matters because the system is not just saying "this well will be better." It is offering an explanation of what changed in the well design and frac execution, and why the production outcome moved with it.
This chart is being used to support the argument that cumulative C-Factor tracks with production performance. In the Halo material, 6-month production showed a statistically significant relationship with cumulative C-Factor. [Confirm final public wording and exact framing.]
The point of this work is to demonstrate that production is predicted accurately by a well's total cumulative C-Factor, which is calculated by summing the final C-Factor for each individual frac stage.
This is the CF-Production curve for our case study field:
Where:
y = 6 month cumulative production,
x = cumulative C-Factor per well.
In Halo's field data, cumulative C-Factor was used to build a production profile and then tested against a held-back reserve well. That combination of correlation plus blind-test validation makes the proof commercially relevant.
Below, the page moves from correlation into why the method works, how landing depth changes the outcome, and how the same signal becomes useful for forecasting and optimization.
We have redefined fracture compliance from its traditional pre-frac injection tests or post-ISIP leak-off to instead measure compliance the way the medical field uses it in blood vessels, a frac's ability to accept a given amount of fluid for a unit of pressure.
This framing matters because it turns compliance from a static after-the-fact concept into something that can be interpreted during pumping, when the operator still has decisions to make.
The medical analogy is useful because it makes the concept intuitive: a more compliant system accepts more fluid for less pressure response, while a less compliant one resists it.
For well 100/15-17, shown in dark blue, 10 stages were fracced into the Doig above the Montney. On its face that is a reasonable completion approach, but the C-Factor profile shows those stages falling between 0.2 and 2.1 while the stronger fracs in the same field score between 3 and 5.
In C-Factor terms, that gap matters because lower values imply the live fracture system is accepting less fluid for the fitted pressure response, which points toward lower effective permeability and smaller productive fracture area relative to the better-performing stages nearby.
For this area, a frac that does not achieve a C-Factor of at least 3.1 is not economical.
This graphic shows AI-predicted C-Factor stage by stage, with each well represented by a different color. It matters because the model is not generating one abstract score per well after the fact. It is producing stage-level output that can be compared against geology, landing depth, pump behavior, and later production.
That makes C-Factor useful not only for forecasting outcomes, but for diagnosing why one interval worked better than another and where the completion design or placement likely reduced value.
This overhead view shows well placement relative to the S2 target. Darker blue indicates intervals that remain within roughly 3 to 6 feet of the target zone, while greener intervals drift farther away, closer to 18 to 20 feet from target.
This stage-by-stage frac profile shows pressure, distance to the S2, and AI-calculated C-Factor moving together through the well, making the landing-depth signal visible during the frac rather than only in postmortem interpretation.
If you look at the northern most well on the over head well map you will see well 103/10-20 and you will notice the green sections of its heat map match the distance to the S2 in this his resolution frac plot
The well profile shows when the borehole porpoises away from the S2 and when it returns. That physical movement aligns with the distance-to-S2 trend seen in the frac profile.
What matters here is not just that the well path moves. It is that C-Factor moves with it. As the wellbore drifts away from the S2, C-Factor falls. As the well returns toward the S2, C-Factor rises.
That behavior is important because it shows C-Factor responding to reservoir quality and pressure conditions in real time while the frac is still underway. In practical terms, it supports the claim that C-Factor is measuring something physically meaningful about permeability, fracture effectiveness, and productive contact with the target interval.
This is why landing depth is not just a geosteering detail. It becomes an optimization variable that can be studied through the frac response itself.
When the frac is landed more than 20 feet from the S2, it becomes separated from the target by a frac barrier. In those stages there is a distinct pressure breakdown mid-frac. Until now, that event was something engineers could observe but not confidently explain.
What matters here is what happens next: C-Factor, shown in purple, rises rapidly after that pressure breakdown occurs. The interpretation is that the frac has broken through the barrier and grown into the S2, and C-Factor is measuring the resulting increase in fracture area and effective permeability immediately.
Now that the event has meaning, and now that a C-Factor KPI of 3.8 has been established for these wells, the frac company can be directed to continue pumping until the stage is fully stimulated. On the right-hand side, where the well is situated closer to the S2, C-Factor starts higher and ends higher because the frac initiates in the S2, and there is no comparable pressure breakdown on those stages.
This graphic highlights the pressure break, proppant response, and AI-calculated C-Factor signal that together show when the frac breaks through into higher-quality rock.
Here we zoom into the ideal section of the well where it is landed within 6 feet of the S2 and notice that three stages have been pumped differently. Two stages grab rate quickly, while the middle stage grabs rate more slowly.
The result is that the slow-rate stage starts with a lower C-Factor and ends with a lower C-Factor, while the stages that grab rate faster start higher and end higher.
We have seen this play out in other fields as well. That matters because it means equipment problems such as a frac blender or pump going down during a stage are no longer unknown, non-quantifiable events. We can now measure the barrels of oil and BCF lost because of frac company equipment issues.
This graphic isolates three stages in favorable geology and shows that differences in rate acquisition alone can still drive materially different C-Factor outcomes.