The Causal Convenience Index

Every quarter, using nothing but public data, Seurat Analytics measures how much of convenience retail's sales movement operators actually control vs. how much was influenced by forces beyond their control.

Last updated July 2026

Inside the store, about half the outcome is yours

We ran thirty years of public data through our model at two levels of resolution: the industry's fuel-inclusive top line, and inside-store sales on their own.

⛽ Fuel-inclusive top line: 96% of the movement comes from forces no operator controls. Most of it is the price of gas, which gets set by global oil markets and is essentially handed to store operators.

🏪 Inside-store sales, sourced from the four public chains that report them separately: 48.5% of the movement in these sales is attributable to outside forces. The other half is made up of the levers operators can pull themselves.

Four chains, the same conditions, very different results

Everyone in convenience faces the same weather, the same economy, the same gas prices. So why do some chains grow faster than others, year after year?

We compared the four public chains against identical national conditions, using only their own reported sales numbers.

📈 Casey's consistently runs 2.9 points above what background conditions predict. Quarter after quarter, for years.

➖ Murphy USA and Couche-Tard's US stores both sit at the line, within the range where we cannot separate them from the conditions themselves.

📉 Arko consistently runs 4.1 points below.

That is a 7.0 point spread between the strongest and the weakest, among operators handed the same environment.

In the quarter ending October 2024, facing the same national conditions, Casey's inside sales grew 4.0% and Couche-Tard's US stores fell 1.6%. A 5.6 point gap, same months, same country, same weather.

External conditions explain when everyone rises and falls together. They do not explain the distance between the chains. Having walked Casey's stores in multiple states and tasted their award-winning pizza and private label foods, it's clear that execution, and crave-worthy, signature items that drive repeat business are making a difference in that retailer's sales.

One honest caveat. The factors we measure are national. A chain's own regions will likely perform better or worse than the country as a whole, and the operator's half will be rooted in different local weather and local economic conditions. Distinguishing local performance from operator skill takes store-level data, which is the work we do directly with operators. Depending on where a chain sits, and local conditions, there may be more than 50% of sales performance available for operators to optimize.

What actually moves the inside of the store

Across the full history of the panel, four outside forces do most of the work on inside-store sales. These are long-run averages, not this quarter's reading.

🥶 Cold weather, 28%.* Colder quarters pull inside sales down. We tested heat separately, and it does not move the store the same way. It is not weather in general. It is cold, specifically. What we cannot tell you from public data is why: households spend more on heating and have less money left over, and people are less willing to get out of the car and walk in. Both are consistent with what we see. Separating them will require store-level data and analysis.

*This is a national model, and national averaging flattens weather. Hot regions and cold regions offset each other in the aggregate. At the level of a single store, weather is likely to rank far higher than 28%.

💼 Unemployment, 24%. A one point rise in unemployment costs about two points of inside sales.

⛽ Pump price, 23%, and it pushes against the store. A 10% rise in gas prices takes about a point off inside sales. Expensive gas inflates the industry's total dollar sales and shrinks the in-store basket at the same time. The money people had in mind for the stop gets eaten at the pump before they reach the door.

🧾 Inflation, 17%. When prices rise, receipt values rise with them. So part of any comp increase is not more items sold. It is the same items sold at a higher sticker price.

What we are not explaining, and why that is the point

Fair question: if outside forces explain about half of inside-store sales, what explains the other half?

Not this model. And that is deliberate.

We only put things in here that no operator controls. Weather. Pump prices. Unemployment. Inflation. There is not a single operator decision anywhere in it. So when half the movement comes back unexplained, that half is not noise or model failure. It is the part that belongs to the people running the stores, and we left it alone on purpose.

We did not leave it entirely unmeasured, though. The 7.0 point spread between Casey's and Arko is sitting inside that half, and it is not weather, because we already took the weather out.

So where would a viral moment fit? A TikTok trend, a new coffee program, an LTO, a remodel: all of them live in the operator's half. None of them are in this model. If someone tells you a viral week moved their quarter, this model cannot confirm it or rule it out. What it can tell you is whether the weather and the pump deserve any of the credit, which is usually the first thing people get wrong.

The useful way to read this whole page: we are not claiming to explain everything. We are measuring the ceiling on what you can fairly blame the world for. At the industry top line, that ceiling is high. Inside your store, it is about half.

Let's dig in to the other half: That other half, measuring the effectiveness of your specific in-store activities, is exactly what we do at Seurat Analytics. If you're interested in learning more about the impact of your actions on sales, down to the store level, get in touch: engage@seuratanalytics.com

Decoding this quarter's headline

Q2 2026. Complete quarter. Census data published through June, retrieved July 16, 2026.

The number the trade press will report is the fuel-inclusive one, and it swings with gas prices. Here is what it actually contains.

Industry sales rose 23.2% in Q2 2026 against Q2 2025. Gas prices account for roughly 88% of that move. Baseline drift, the economy's normal upward pull, accounts for about 8%. Every other outside force we measure, weather and unemployment and inflation and the rest, adds up to about 1%.

Everything else, operators included, comes to roughly 2%.

Two plain ways to use this:

✅ Before you celebrate a boom quarter, check the gas prices. Nearly nine tenths of this spring's industry growth was the pump, not performance.

✅ Before you punish a weak quarter, check gas prices, too. In spring 2025 industry sales fell 5%, but conditions alone were worth a 6.4 percentage point drop. Operators slightly beat a bad year and mostly got blamed for it.

Gas prices and store readiness

The Index measures the past. Its inputs do not. Gas prices, unemployment, and the weather are all visible a quarter out. Which means you can know whether the retail tide is about to rise before it actually does.

Here is why that is worth knowing: When gas gets cheaper, the money that a consumer would have spent on gas is still in their pocket when they walk through the door into your store. That money, that "budget" they had for the "gas trip," can now go to whatever you put in front of them.

And it is not a traffic story. We account for how much America is driving separately, and the lift is still there without it. This is the same consumer, making the same stop, just with more room in their budget to spend in store.

So a falling-gas quarter is not a traffic problem. It is a merchandising opportunity. They are already coming. The question is whether what's on your shelf is ready for the newfound willingness to spend.

One more signal worth watching in this section each quarter: when all four public chains beat the conditions at the same time, that is not weather. That is the convenience channel itself winning.

How this is built

Both layers use the same approach: year-over-year sales changes measured against eight forces no operator controls, including pump price, heating and cooling weather, precipitation, unemployment, inflation, miles driven, and consumer mood. The industry layer uses Census Bureau retail sales for US gasoline stations back to 1993. The inside layer uses 94 quarters of same-store results (2019 through Q1 2026) from Casey's, Murphy USA, Couche-Tard, and Arko public filings. Every source is public and every number can be checked.

Census gasoline-station sales are an advance estimate and get revised. Every figure on this page is stated as of the data vintage above. When a previously published number moves, we say so.

Run the same logic on your own numbers

Want the store-by-store version, the one that separates local conditions from what your teams are doing? That is the work we do for clients.

See the full Board → · See how it works → · Detection Floor Check →

Updated quarterly from public data. Q2 2026 is complete. Next update: November 2026, when Q3 closes.