Every QSR brand here is understood, and AI recommendation is where the field breaks apart.

ARA read 15 QSR brands against four AI models and 16 buyer questions. That gives each brand 64 recommendation moments. The median score is 66. Chick-fil-A leads at 87. Pizza Hut sits last at 55.

Clarity is settled. No brand in the set scores below 16 on Clarity. Nine of the fifteen score 18. The models can explain every one of these brands.

AI Recommendation is not settled. Three brands hold the maximum of 20. Jack in the Box and Sonic are named in no answer at all, at any position. Explanation and selection have come apart.

Six of the fifteen brands carry real equity the models under-reward. Two are recommended above their equity. Seven show agreement between the two. Selection is where the remaining gain sits.

ABOUT THIS READING

ARA scores five attributes, each out of 20: Findability, Clarity, Sentiment, Identity Fidelity, AI Recommendation. The five sum to the total, and the total sets the tier. AWESOME is 83 to 100 / STRONG is 70 to 82 / AVERAGE is 56 to 69 / WEAK is 40 to 55 / INVISIBLE is 0 to 39. ARA measured movement on the same 16 questions and the same 15 brands as last quarter, one run per question per model, across four models. AI Recommendation uses the same formula in both quarters. Sonic's rise from 45 to 57 is mostly a corrected measurement rather than brand movement. Last quarter's prompt said only "Sonic", and 17 of its 64 answers described the video game character, so its findability and sentiment scores were partly about the game. This quarter's prompt names the drive-in. Sonic's recommendation reading never says the brand name, so that reading is clean, and it stayed at zero in both quarters.

Two measurements also changed. ARA scored Identity Fidelity on a revised rubric this quarter with five criteria. The revision removed Brand Character and added Brand Messaging. Identity Fidelity movement is therefore not like-for-like, and no reader should treat it as brand movement on its own. Findability now includes a measured agent-readiness check of the brand's own site. Last quarter ARA did not measure that sub-criterion and it defaulted to zero for every brand, so part of any findability rise is the measurement arriving. McDonald's, Panda Express and Shake Shack returned an HTTP 403 to the scanner. The rubric scores a 403 as zero. That capped McDonald's official-source score and pulled its Findability from 14 to 10. KFC and Pizza Hut rate-limited the scanner twice, so ARA records them as not measured on that sub-criterion and their Findability is the sum of four criteria.

The Winners.

Chick-fil-A holds the top at 87, up two points since last quarter. Findability rose one point and Clarity fell one point. The new agent-readiness check explains part of that findability rise. Chick-fil-A is the first choice in 15 of the 64 moments, more than any other brand, and it appears in 47 of them. All four models name it. It wins the most delicious question, the first visit question, the families question, breakfast and cleanest. It is absent from plant-based and from travelling abroad.

Sonic posts the biggest rise, up 12 points to 57. Findability rose from 6 to 13 and Sentiment from 9 to 13. Most of that is a corrected measurement. Last quarter the prompt said only "Sonic", and 17 of the 64 answers came back about the video game character. This quarter the prompt names the drive-in. Sonic is still chosen zero times, and no model names it at any position.

Three bets stood against Sonic's name last quarter and none of them landed. The financial-narrative bet targeted official sources and retrievability. ARA grades it "not acted on", because models still find no current disclosure and do not know the brand president appointed for 2026. Findability rose seven points, and the correction explains that rise, so the bet cannot claim it. The recommendation bet targeted AI Recommendation and ARA grades it "still zero". That reading never says the brand name, so it is the one clean comparison against last quarter, and it is flat at zero. Sonic shipped four product launches in the quarter and none reached the layer that routes a recommendation. The verbal-identity bet targeted Identity Fidelity and ARA grades it "no clean prior", because the rubric changed and the prior reading was partly about the video game character.

The Losers.

Pizza Hut sits at the bottom at 55, down two points, and it is the only weak brand in the set. Findability fell one point and Sentiment rose one point. Pizza Hut rate-limited the scanner twice, so its findability is the sum of four criteria and that one-point fall is not a clean read. One model names it, in 2 of the 64 moments. It wins nothing. It is absent from america's favourite, most delicious, best value and families with kids.

Domino's takes the biggest fall, down three points to 63. Identity Fidelity fell from 15 to 13 and Findability from 14 to 13. The revised identity rubric covers two of those three points, so most of the fall cannot be read as brand movement. Two of the four models name Domino's, in 5 of the 64 moments. Its best showing is the app and rewards question, where it is named but never first. It is absent from most delicious, families with kids, eating healthier and plant-based.

Three bets stood against Domino's name and the record is mixed. The occasion bet targeted AI Recommendation, and that reading rose one point, so ARA grades it "inched up". No occasion content appeared and the site still carries no structured markup, so the lane is open. The food-quality bet targeted Sentiment, and ARA grades it "shipped, slipped". Domino's shipped the flagship platform, no model returned it, and what the public carries fell one point. The technology bet targeted Clarity, and ARA grades it "gap widened". Models still recite the tracker and the ordering platform as trivia, clarity did not move, and the chief executive succession is absent from every read. The tie to the bets stops at Identity Fidelity. No bet targeted it, it caused most of the drop, and the rubric changed underneath it.

The Leaderboard.

Sixty-eight brands share the QSR shelf, and the fifteen measured names hold 58.6% of it.

The shelf counts 326 choices across 68 brands. The fifteen measured brands hold 58.6% of those choices. The largest single holding is 45 choices, or 13.8% of the shelf. The rest of the shelf belongs to names no client asked ARA to score.

The biggest name outside the measured set is In-N-Out Burger with 18 choices. Culver's takes 12. Jersey Mike's takes 9. Starbucks takes 7. Raising Cane's and Whataburger take 6 each. Five Guys and Panera Bread take 5 each. The models put these names on the same shelf as the national chains.

The tail is long and thin. Thirty-five brands are named exactly once, which is 11% of all choices. Five brands are named exactly twice, which is 3%. Together 40 of the 68 brands hold 14% of the shelf. A brand named once reads as noise in the answer set. A brand named 18 times is a competitor.

16 buyer questions x 4 models / audited brands in colour, every other brand in a stepped grey. Toggle between Cohort Only and Full Shelf.

KEY FINDING 01

No brand in the set scores below 16 on clarity, and the totals still run from 87 to 55.

Clarity runs from 16 to 19 across the fifteen brands. McDonald's holds the highest at 19. Nine brands score 18. The two lowest, Panda Express and Pizza Hut, score 16. That is a narrow band.

AI Recommendation runs from 20 to 0 across the same fifteen brands. The two attributes describe different problems. One is close to solved for everybody. The other separates the leader from the last brand on the board.

Subway sits exactly at the median, 66. A brand in that position cannot buy a better rank with more explanation. The models already explain it clearly. The gain has to come from being chosen.

KEY FINDING 02

Jack in the Box matches Burger King on identity fidelity, and no model named it once.

Jack in the Box scores 13 on Findability, 18 on Clarity, 13 on Sentiment and 17 on Identity Fidelity. Burger King and KFC also score 17 on Identity Fidelity, and both sit in the strong tier. Jack in the Box scores 0 on AI Recommendation. No answer names it, at any position.

Sonic reads the same way at zero. Arby's scores 16 on Identity Fidelity and appears once. Pizza Hut appears twice. Shake Shack scores 18 on Clarity and 15 on Sentiment and appears five times. Good fundamentals do not produce a recommendation on their own.

Six brands hold a swell position in this reading: Domino's, Arby's, Jack in the Box, Shake Shack, Sonic and Pizza Hut. Their equity is real and the models under-reward it. That is the cheapest gap on the board to close, because the brand work already exists.

Brand Fundamentals vs Recommendation.

KEY FINDING 03

Subway scores 17 on clarity and 11 on identity fidelity, the lowest identity score in the set.

Every model can explain what Subway sells. None of them returns a strong sense of who Subway is. Shake Shack and Domino's show the same shape at 18 on Clarity and 13 on Identity Fidelity. Pizza Hut scores 16 and 12. Panda Express scores 16 and 13.

Chick-fil-A and Wendy's score 18 on both. Taco Bell, McDonald's, Burger King, KFC and Jack in the Box score 17 on Identity Fidelity. In this reading the five lowest identity scores all belong to brands in the average or weak tiers.

One caution holds here. ARA scored Identity Fidelity on a revised rubric this quarter, so these levels compare across brands in this reading and not against last quarter. The pattern inside the quarter still stands: being understood is common, and being distinct is not.

Understood vs Distinct.

KEY FINDING 04

Three brands score the maximum on AI recommendation, and Chick-fil-A leads all three on first place.

Chick-fil-A, Taco Bell and McDonald's all score 20 on AI Recommendation. Their first-place counts differ. Chick-fil-A is first in 15 of the 64 moments. McDonald's is first in 11. Taco Bell is first in 10. Their presence differs too, at 47, 34 and 32 moments.

Chipotle scores 19, is first in 5 moments and appears in 20. The recommendation score reaches its ceiling before primacy does. A maximum score says a brand is present almost everywhere it should be. It does not say the models lead with that brand.

The lower half shows the same care is needed. Panda Express appears in only 4 moments and is still first in 1 of them. Wendy's scores 17, appears in 18 moments and is first in 2. Any claim of leadership has to cite the first-place count.

KEY FINDING 05

McDonald's is chosen more often than its retrievability supports.

McDonald's scores 80 and sits in the strong tier. Its Clarity is 19, the highest in the set. Its AI Recommendation is 20, it is first in 11 moments and it appears in 34. Its Findability is 10, the lowest in the set with Panda Express. ARA reads that as an ebb position.

Part of the findability number is the measurement. The McDonald's site returned an HTTP 403 to the agent-readiness scanner, the rubric scores a 403 as zero, and that capped the official-source score and pulled Findability from 14 to 10. The 403 is also a controllable input. A site that answers the scanner scores the sub-criterion it was built to score.

Subway is the other ebb brand. It scores 66, the median, with Identity Fidelity at 11 and AI Recommendation at 15. An ebb position carries risk in both cases. If what the models hold about the brand drifts, the recommendation may follow it down.

WHAT THIS MEANS

Clarity is finished work in QSR. Nine brands score 18 and none falls below 16. The spread from 87 to 55 comes from selection instead. Six brands hold equity the models under-reward, and two are recommended above theirs. The next point of movement sits in the recommendation layer, and nowhere else on the board is it this cheap to take.

The bet record is the warning. Sonic shipped four launches and stayed at zero. Domino's shipped its flagship platform and no model returned it. Both brands published product news, and neither reached the layer that routes a recommendation. Chick-fil-A leads the set at 87 and is still absent from plant-based and from travelling abroad. A brand that wants a different answer next quarter has to publish something a model can retrieve and compare.