Your Comps Are Up, and Your Traffic Is Down. Which One Is the Business?

Wendy's raised its average check 5.6% in Q2 2026 and served 12.5% fewer guests. U.S. same-restaurant sales landed at -7%. Cava posted +9.0% the same quarter, 5.3 points of it from guest traffic. Both numbers get called comps. Only one of them is a business.

The word hides the difference. A chain can charge more, serve fewer people, and still print a positive comparable sales number for several quarters. Then the math runs out.

Wendy's is what running out looks like.

Split the comp into three parts and read the traffic line alone. Price tells you what you charged. Mix tells you what guests added. Traffic tells you how many humans chose you. Only the third one is demand.

Cava publishes its split. Q1 2026: same-restaurant sales +9.7%, guest traffic +6.8 points, menu price and product mix +2.9 points. Q2 2026: +9.0%, traffic +5.3, price/mix +3.7. Average unit volume climbed to $3.1 million from $2.9 million.

Note the label. Cava reports "price and mix" as one line, not price alone. Anyone who tells you Cava raised prices 3.7% is reading the filing wrong.

Wendy's split runs the other way, and its own CFO said so out loud. Steven Cirulis, Chief Financial Officer and Chief Strategy Officer, told the Q2 call: "The decline in U.S. same-restaurant sales was driven by a 12.5% decrease in traffic." He noted it was "partially offset by a 5.6% increase in average check." The company's filed results put U.S. same-restaurant sales down 7.0%.

Read both sets as a report and you see a strong quarter and a weak one. Read them as a dining room and you see something else. One restaurant has more people in it than last year. The other has a shift manager watching the 12:30 rush thin out. She cuts hours. Then she gets asked on a call why the value proposition isn't landing.

Robert Wright, President and Chief Executive Officer, said it without decoration on the same call: "Our quality differentiation has eroded." And: "Our value proposition has weakened." Those are traffic statements. No amount of check growth answers them.

Three inputs, one period, one prior-year comparison. Guest count. Net sales. Item-level menu prices. The rest is arithmetic you can do in a spreadsheet.

Start with guest counts, not dollars. Pull transaction count for the comp period and the same period last year. The change between them is your traffic line. It is the only number a price sticker cannot inflate.

Then average check. Net sales divided by transactions, this year against last. That delta is price and mix combined.

Then pull them apart. Price is what you changed on the menu board, weighted by unit volume. Mix is the remainder. Upsells, drink attach, the premium protein, the guest who added chips. If check rose 5% and you took 2% of pricing, mix did the other 3%. Worth knowing. Mix reverses without warning.

Here is one location, real shape, round numbers. March: 9,400 transactions, $221,000 net sales, $23.51 average check. Last March: 10,200 transactions, $214,000, $20.98. Comp reads +3.3%. Check reads +12.1%. Traffic reads -7.8%.

The P&L says growth. The dining room lost 800 visits.

Most POS systems hand you transaction counts in two clicks. Most operators never pull them. They watch the sales line, because the sales line is what the bank asks about.

Less than you think, and the number is specific to your guest. Cava took roughly 1.4% in January 2026 and grew traffic 6.8 points that quarter. Study that ratio.

Brett Schulman, Co-Founder and Chief Executive Officer of Cava, described the move on the Q1 call: "This past January, we took an approximate 1.4% price increase while holding bowl and pita pricing flat." He added that Cava's price adjustments since the end of 2019 run "only slightly more than half of cumulative CPI."

Read that again. A chain compounding comps near 9% has priced at half of inflation for six years. That is not a pricing strategy. That is a decision to stay cheaper than the number the guest carries around in their head.

Guests are doing that math. Circana found 54% of U.S. households under $75,000 in income said better prices would raise their restaurant visits. Value-menu traffic rose 1% while overall chain traffic fell 1%. Both figures sit in L.E.K. Consulting's 2026 restaurant playbook.

Your own price elasticity is already in your POS. Take every item where you moved price in the last year. Compare unit volume eight weeks before against eight weeks after. Do ten items. Two or three will show volume falling faster than the price gained. Those are the ones costing you money.

The industry version of this exercise involves a consultant and a deck. The useful version takes an afternoon.

Order count and average order value, tracked separately, every week, per brand. A virtual brand can post rising delivery revenue while losing customers. No marketplace dashboard will flag it for you.

The structure matches a dining room with one difference. You cannot see the people. A physical restaurant signals trouble when the lunch rush looks thin. A delivery-only brand signals nothing until the report lands.

So the split matters more here, not less. Order count → demand. Average order value → price and mix. Revenue → the two multiplied, and the one number that lies.

Pull it per brand and per platform. A kitchen running four virtual concepts has four demand curves, and they rarely move together. One brand can carry the revenue line while the other three shed orders. The blended figure shows growth. Three of your four concepts are dying inside it.

L.E.K. puts close to 75% of restaurant traffic off-premises now. That is a lot of demand nobody watches in person.

Guest counts, weekly, against last year, on one chart, visible to whoever runs the shift. Not monthly. Not in a deck.

The industry is swinging back toward traffic whether operators are ready or not. Black Box Intelligence logged December 2025 comps at -1.0% with traffic at -3.3%. January 2026 flipped comps positive at +1.0%, traffic still negative at -1.1%. The gap between those two lines is pricing. It has been carrying the industry for three years.

L.E.K. Consulting says the carrying is over. Its playbook comes from Alison Schilling and Chris Randall, both Managing Directors and Partners at the firm, with the operator view from Abigail Pringle, former President of the U.S. business at The Wendy's Company. A former Wendy's executive signing the case against price-led growth reads differently after this quarter.

Cava raised full-year same-restaurant sales guidance to 4.5%–6.5% from 3%–5%, and that raise rests on guest counts. Pull your own transaction count for last week before you read another industry forecast.