A shale well tells you most of what it's going to be within its first 90 days. By then the completion has cleaned up, the frac water is largely back, and the well is producing on its own reservoir energy. Wait a year and you'll have a cleaner decline curve — but you'll also have made every decision that mattered. Capital allocators, A&D teams, and completions engineers all want the same thing: a defensible read on well quality as early as the record allows.
That read starts with early-time production — the IP30 (average daily rate over the first 30 producing days) and IP90. Neither is a perfect predictor of EUR, but both are the first hard numbers you can benchmark, and they separate the top-quartile wells from the pack faster than anything else on the record.
The question a user actually asks
"Of the twelve wells this operator brought online in the county over the last year, which ones are outperforming — and by how much — before the decline even sets in?"
That's a question you can answer against the Wellsite data lake in plain language. The platform pulls each well's production history, isolates the first 30 and 90 days of reported volumes, and lines them up side by side. No spreadsheet, no manually stitching monthly allocations to first-production dates.
Normalize before you compare
Raw IP30 is a trap. A well that flowed 28 of its first 30 days will look worse than one that reported a partial first month, unless you account for producing days. So the first move is to convert to a rate — barrels of oil per day and mcf per day over actual on-production days — rather than comparing raw monthly cumulative volumes.
From there, the comparisons that matter:
- Against the county average. How does this well's IP90 stack up against the typical well completed in the same county? This is the coarsest benchmark, but it frames everything.
- Against direct offsets. Wells in the same section or immediately adjacent share rock. If a new completion beats its offsets by 40% on IP90, that's a completion or landing-zone story worth understanding.
- Against the operator's own book. Is this the operator's best early-time well in the county, or a middle-of-the-pack result dressed up in a press release?
The Wellsite data lake handles all three framings from the same starting question, because it already carries the county history, the offset wellbores, and the operator's full record.
What early rate does — and doesn't — tell you
A strong IP90 confirms the reservoir is there and the completion connected to it. That's real information. But early rate can be inflated by an aggressive choke, and a well flowing wide open early may simply be pulling its future forward. Two wells with identical IP90s can diverge hard by month twelve if one was managed on a tighter choke.
So pair the early rate with the shape of the first few months. A well holding flat through its first 90 days is telling a different story than one already down 30% month over month. The record shows both the level and the slope, and the slope is where the choke-versus-decline question starts to get answered — the same distinction that separates a genuinely fading well from a choked-back one.
Outlier detection matters here too. A well that reports an enormous first month followed by a collapse often reflects flush production or a reporting artifact, not sustainable deliverability. Flagging those early keeps them from skewing a type curve or an acquisition model.
Turning it into a screen
Once you've got a normalization method you trust, the same logic scales into a screen. Rank every well brought online in a county over the last twelve months by IP90, filter to a single operator or formation, and you have a shortlist of the best recent completions — and the laggards — in minutes. For an investor sizing up an operator's recent program, that ranking is the difference between taking a completion narrative at face value and checking it against the record.
For an operator, the same screen turns into a feedback loop: which of your recent designs are producing the strongest early rates, and are your newest wells beating your older ones? That's vintage analysis you can run before the wells have enough history for a full decline fit.
The honest caveat
Early production is a leading indicator, not a verdict. A great IP90 on a well that was drained too fast can still disappoint on cumulative recovery, and a modest early rate on a tightly managed well can outlast its flashier neighbors. Use the first 90 days to sort and prioritize — then let the decline curve confirm or correct the story as the months accumulate.
The point is speed. The record gives you a benchmarked, offset-aware read on well quality months before a clean decline fit is possible. When you're deciding where to deploy the next rig or whether an operator's recent book justifies its ask, months of lead time is worth a lot.