Dimension Labs · Analysis · Core-category read
2024-25 season · seven structured survey categories · 117,975 responses · 40,111 linked comments
This report — both its analysis and its design — was produced by the Dimension Labs AI Agent (Beta) as a one-shot response to a single query, “identify what mattered most to fans in the 2024-25 season based on survey feedback across the league,” with every finding independently validated.
Bottom line
Fans rate staff and the in-game show well. They punish concession and parking prices, digital reliability, and problems that never get resolved. Every driver below is an aspect inside a structured-survey category — controllable, and pointed straight at NPS and overall satisfaction (OSAT). The cross-cutting “cost & value” theme that ties several of these together was derived from the raw text by AI, not asked in the survey; it is covered in the Derived categories section at the end.
Per the analysis plan, the spine is the structured-survey categories and the drivers are the aspect-level subcategories inside them. Each move is controllable at the club level and grounded in the evidence that follows.
| Survey category | The aspect driving it (proof) | Move that raises the score |
|---|---|---|
| 1Food & beverage | Price 73.9% bottom-two-box (n=1,990) · sold-out items at a 100% issue rate · largest category at 5,700 comments | Fix price architecture, stockout discipline, and checkout speed before touching the menu. |
| 2Transportation | Parking cost 85.1% bottom-two-box (n=522) · arrival traffic 77.9% on 2,553 comments | Split into parking economics, lot flow, and wayfinding — three owners, three fixes. |
| 3Venue technology | App reliability 6.86 NPS & 94.2% bottom-two-box — the lowest-rated survey category (2.41) | Run a reliability audit of the app, mobile tickets, and the scan path before adding features. |
| 4Staff | Issue resolution 7.12 NPS vs 9.39 on friendliness — the category is a strength, this aspect isn’t | Protect the friendliness culture; tighten escalation, handoff, and exception handling. |
Control
Winning helps. It does not fix the categories.
NPS climbs from 8.48 after losses to 9.17 after wins, and top-box OSAT from 46.7% to 56.2%. But the category-level pain in food, parking, and technology persists across wins and losses alike — these are operations problems, not scoreboard problems.
Rank the seven structured-survey experience areas and the picture is concentrated: three underperform, three are healthy, and staff is a clear strength. The action sits in the aspects inside the weak ones.
Average NPS among fans who raised each aspect, within its linked mention base. Every aspect shown belongs to a structured-survey category. The dashed line marks the all-response average of 8.84.
Each aspect belongs to a survey category (shown at right): app reliability, ticket scanning, and mobile ticket access to venue technology; issue resolution to staff; sold-out food and concessions price to food & beverage; parking cost to transportation. The two deepest holes are low-volume but severe; concessions price and parking cost sit closer to average but carry far more volume.
avg_nps within each linked aspect base · 2024-25 season · aspect n shown per row · Dimension Labs enrichment layer
Evidence. Ranked by bottom-two-box dissatisfaction, the seven survey categories fall into three bands. Venue technology (49.0%, and the lowest average rating at 2.41), transportation (47.4%), and food & beverage (39.5%) underperform. Merchandise (25.0%), entering & exiting (26.8%), and in-game entertainment (18.4%) sit in the middle. Staff (12.6%) is the clear strength. But the actionable drivers are one level down: the aspects inside those categories — parking cost, app reliability, concession price, issue resolution — are where the promoter-score damage concentrates.
Interpretation. This is the plan’s framework in one view: hold the primary categories as the accountable spine, then diagnose at the aspect level, because operators act on aspects, not abstract categories. The fastest route to higher NPS and OSAT is to work the weak aspects inside the three underperforming survey categories, in order of how deeply each drags the score.
Qualification. These are salience signals from fans who chose to write comments, not attendee incidence rates, and each category carries its own linked base.
| Survey category (the accountable spine) | Base | Avg rating | Bottom-2-box | Issue rate |
|---|---|---|---|---|
| Venue technology | 575 | 2.41 | 49.0% | 84.1% |
| Transportation | 2,553 | 3.25 | 47.4% | 61.4% |
| Food and beverage | 5,700 | 3.24 | 39.5% | 64.8% |
| Entering & exiting | 2,419 | 3.97 | 26.8% | 36.9% |
| Merchandise | 1,414 | 3.53 | 25.0% | 67.7% |
| In-game entertainment | 3,136 | 4.17 | 18.4% | 37.2% |
| Staff | 5,369 | 4.45 | 12.6% | 21.0% |
Concessions is the biggest linked survey category and the most exposed to a worth-the-price breakdown. Inside it, price dominates — with checkout and stockouts as the fastest fixes.
Bottom-two-box rate for each food & beverage aspect, sorted worst-first. The dashed line is the category average of 39.5%.
Longer bars are worse. Price towers over every other aspect and carries by far the most volume — but sold-out items and clunky checkout are smaller, sharper problems a club can fix faster.
bottom-two-box within each aspect’s linked mention base · food_beverage aspects · 2024-25 · n per row
Evidence. Food & beverage is the largest linked survey category at 5,700 comments, with a 3.24 average rating and a 64.8% issue rate. Inside it, price dominates: 1,990 mentions, a 2.12 average rating, and 73.9% bottom-two-box. Ordering and checkout is smaller but sharp at 46.4%, and sold-out inventory is the harshest discrete failure — a 1.88 rating and a 100% issue rate on 185 mentions.
Interpretation. The first food & beverage question is not “how do we improve the menu.” It is “where are fans paying too much for what they receive.” That points to price architecture, bundle design, inventory discipline, and checkout speed — and it is the largest single volume of NPS-dragging comments in the study.
Qualification. This is a text-derived read of the food & beverage survey category. It reflects what comment-writers chose to raise.
Have more affordable concessions. We paid double what I would expect to pay for 2 burgers, cheese curds, and a few cokes.
The transportation question explicitly covers parking and transit. Fans answer it as a cost-and-flow problem — parking cost, lot flow, and arrival congestion they remember as part of the game.
Bottom-two-box rate by transportation aspect, worst-first. The dashed line is the category average of 47.4%.
Every aspect sits above the 47.4% category line — there is no bright spot here. Parking cost leads by a wide margin, with arrival traffic and lot flow close behind.
bottom-two-box within each aspect’s linked mention base · transportation aspects · 2024-25 · n per row
Evidence. Transportation carries a 3.25 average rating and a 61.4% issue rate on 2,553 comments. Parking cost is the largest complaint at 522 mentions, a 2.23 rating, and 85.1% bottom-two-box. Parking-lot flow adds 311 mentions at 72.3%, and arrival traffic runs a 2.10 rating and 77.9% on 222 mentions.
Interpretation. Transportation should be split into parking economics, pregame throughput, and navigational clarity. They are related failures that need different owners and different fixes — a club can improve lot flow without touching the price fans resent, or vice versa.
Qualification. Team-level transportation coverage is thinner than food & beverage, so finer rank differences should stay descriptive.
Parking took a long time and was more expensive than advertised online.
Technology complaints are fewer than concessions complaints, but they are sharper. When the app or ticket path fails, fans feel stuck — and it shows in the lowest promoter scores in the study.
Bottom-two-box rate by venue-technology aspect, worst-first. The dashed line is the category average of 49.0%.
App reliability nearly maxes the scale — almost every fan who raised it was dissatisfied. Mobile ordering, by contrast, mostly works; the problem is concentrated in access, not features.
bottom-two-box within each aspect’s linked mention base · venue_technology aspects · 2024-25 · smaller-base category, n per row
Evidence. Venue technology is the weakest-rated survey category at a 2.41 average rating, 49.0% bottom-two-box, and 84.1% issue rate on 575 comments. App reliability is the clearest failure — 69 mentions, a 1.58 rating, 94.2% bottom-two-box, and a 6.86 NPS, the lowest of any aspect in the study. Mobile ticket access is also weak (2.15 rating, 55.1%), while mobile ordering is comparatively healthy.
Interpretation. The priority is not a generic technology upgrade. It is a reliability audit of app access, mobile tickets, and the scan path — the workflows fans hit at the door and at their seat. Because these failures produce the deepest NPS holes, fixing a small number of blocking bugs returns outsized promoter-score gains per fan touched.
Qualification. This is the smallest survey-category base in the study, so team-by-team separation inside technology is diagnostic rather than conclusive.
Courtesy and helpfulness are clear assets and the highest-rated survey category. The weak point is a single aspect: what happens when something has already gone wrong.
Bottom-two-box rate by staff aspect (lower is better). Green marks the warmth strengths; red marks the resolution break point. The dashed line is the category average of 12.6%.
Dots to the left are strengths; dots to the right are weaknesses. Friendliness and helpfulness sit far left — issue resolution sits far right, more than thirteen times the dissatisfaction of friendliness.
bottom-two-box within each aspect’s linked mention base · staff aspects · 2024-25 · n per row
Evidence. Staff is the strongest survey category at a 4.45 average rating and 12.6% bottom-two-box on 5,369 comments. Friendliness runs at just 3.6% (9.39 NPS) and helpfulness at 7.7% (9.26 NPS). The break point is issue resolution: 49.8% bottom-two-box, an 80.9% issue rate, and a 7.12 NPS on 209 mentions — the second-lowest promoter score in the study.
Interpretation. The staff agenda is not culture repair. It is escalation, handoff, and follow-through once something has already gone wrong. Because a fan whose problem goes unresolved scores more than two full NPS points below one who is simply greeted well, the fastest OSAT gain here is a working exception path — not more training on warmth the league already has.
Qualification. Categories and aspects can co-occur, so a staff complaint can also sit inside entry, concessions, or transportation frustration.
Three survey categories round out the seven. Two are relative strengths; the third carries a quieter warning worth a closer look next run.
In-game entertainment and entry are working. In-game entertainment posts the second-lowest dissatisfaction of any survey category — 18.4% bottom-two-box and a 4.17 average rating on 3,136 comments — so the halftime-and-timeout product is a genuine asset, not a fix surface. Entering & exiting the arena sits close behind at 26.8% bottom-two-box and a 3.97 rating on 2,419 comments: solid, with entry friction concentrated in peaks rather than a systemic failure.
Merchandise is the one to watch. Its dissatisfaction looks mild — 25.0% bottom-two-box, a 3.53 rating — but it carries a 67.7% issue rate on 1,414 comments, the second-highest issue rate of any survey category behind venue technology. That gap between a modest bottom-two-box and a high issue rate says fans are flagging specific friction (selection, checkout, price) more than they are rating the category poorly overall.
What we are not claiming. Merchandise, entry, and in-game entertainment have no persisted aspect-level breakdown in this run, so we do not name a driver inside them — the plan is explicit that a driver requires aspect-level evidence, not category volume alone. Merchandise incidence is also a known reconciliation item. The recommendation is narrow: add a merchandise aspect pass in the next run to convert that 67.7% issue rate into a nameable, fixable cause.
Benchmarking belongs on the categories the survey asks. Transportation is the clearest primary-category team signal — some clubs carry sharper pain, but the right reading pairs the rate with the base.
Transportation bottom-two-box rate by club, with linked base beside each bar. The dashed line is the league transportation average of 47.4%.
San Antonio and Atlanta show the sharpest transportation dissatisfaction, but on ~70–150 comments each. Dallas’s 53.3% is more moderate — yet it rests on 1,349 comments, the only club base deep enough to act on with confidence.
transportation bottom-two-box by team · minimum 50 linked comments · 2024-25 · descriptive comparison only
Evidence. On transportation, San Antonio (83.9% bottom-two-box, n=149) and Atlanta (83.8%, n=68) post the sharpest dissatisfaction, followed by Oklahoma City (60.0%, n=50). Philadelphia (54.3%, n=105) and Dallas (53.3%, n=1,349) are more moderate — but Dallas’s base is an order of magnitude deeper than any other club’s.
Interpretation. Two readings, two actions. San Antonio and Atlanta are candidates for a targeted transportation diagnostic where the rate is extreme, treated as directional given the base. Dallas is where a league-versus-club comparison is statistically safe — a moderate rate that holds across nearly 1,400 comments is a firmer basis for club-level work than a sharp rate on 68.
Qualification. These comparisons are descriptive only; thin-base clubs (Atlanta, Oklahoma City) are directional flags, not decisive rankings. A full testing layer was not persisted across every club cut.
The result raises the ceiling on how fans feel, but comments about the survey categories still score below comments about the result itself — evidence the experience tax is real and separate from the scoreboard.
Average NPS by comment type in the win/loss layer. The shaded band spans the loss-night (8.48) to win-night (9.17) average — the swing the result alone produces.
The band is the mood swing a win or loss creates. Food and transportation comments fall inside it — below comments that mention the result (9.25). Even on the same night, the categories fans raise pull the score down relative to the game itself.
avg_nps in the win/loss context layer · n=43,732 linked · win/loss averages on 60,175 / 57,800 responses · directional, not causal
Evidence. Structured NPS averages 9.17 after wins and 8.48 after losses, and top-box OSAT runs 56.2% versus 46.7%. In the linked layer, comments about the survey categories score below comments that mention the result: transportation 8.82 and food & beverage 8.89, versus 9.25 when the result itself is raised.
Interpretation. Winning lifts the mood, but operational categories still score below the result on the same night — so weak experience feedback cannot be dismissed as a bad-night effect, and a fixed category story can raise season-long NPS independent of the standings. The cross-cutting cost & value lens sharpens this further (it scores below even the loss-night average) — see the Derived categories section.
Qualification. This is a directional operations-versus-result read, not causal proof.
Buyer type is a structured survey field. Members and multi-game buyers rate the recurring survey categories more harshly and score lower on NPS — which ties this work to retention.
Food & beverage bottom-two-box rate and average NPS by buyer type. Members and multi-game buyers are highlighted.
On food & beverage the darker (member) bar runs longer — more dissatisfaction — and members’ average NPS sits nearly four-tenths of a point lower. The split is in the recurring basics fans pay for repeatedly, not in perks.
buyer_type × food_beverage bottom-two-box and structured nps · single-game n=76,217 · member n=39,338 · directional
Evidence. Single-game buyers average 8.96 NPS on 76,217 responses; members and multi-game buyers average 8.57 on 39,338. On food & beverage, single-game buyers run 33.7% bottom-two-box versus 52.9% for repeat buyers — the split is in the recurring basics fans pay for again and again, not in premium access.
Interpretation. Repeat buyers are more sensitive to recurring survey-category breakdowns than to premium shortfalls, which makes this work a retention lever, not just a satisfaction one — the fans with the most future revenue attached are the ones the basics are losing. The derived cost & value lens shows the same split even more starkly (see the end).
Qualification. This is directional only. No downstream attendance or spend outcomes were available for predictive fan-value modeling.
The consession prices are up 20% and the food is not better. I will no longer eat at the games. It's not worth the money.
Four moves, each an aspect inside a survey category, ordered by payoff. Controllable, grounded in the evidence above, and startable this off-season.
Price is the dominant aspect in the largest category. Start with price architecture and bundles, stockout discipline, and checkout speed — before any menu redesign.
73.9% B2B on price; 5,700 comments · Owner: F&B / vendor management
Parking cost, lot flow, and arrival traffic are distinct aspects with distinct owners. Combining them into one transportation bucket blurs accountability.
85.1% B2B on parking cost · Owner: venue operations & parking
The weakest technology aspects happen at critical access moments. Reliability beats novelty; a handful of blocking bugs produce the deepest promoter-score holes.
6.86 NPS on app reliability · Owner: digital product / ticketing
Warmth is working; issue resolution is not. The staff agenda is follow-through, not culture repair — a working exception path returns outsized OSAT per fan touched.
7.12 vs 9.39 NPS gap · Owner: guest services / venue ops
Derived categories
Everything above comes from the seven experience questions the survey actually asks. The categories in this section were not on the questionnaire.
They were derived directly from the raw open-text — the “how can we improve” and “what went well” answers — by an AI enrichment pass that reads every comment and tags cross-cutting themes no fixed question captures. Seven survey questions produced fourteen enriched categories.
This is where a single dataset becomes extensible. Cost & value, venue comfort, game presentation, ticketing, premium access — none required a new survey item, a new field, or a new mailing. The league can pose a research question after the data is collected — “is this really a value problem?”, “is the ticketing flow costing us?”, “how does comfort read?” — and answer it from feedback fans already gave. The same comment about $18 nachos is at once a food & beverage price signal and a cost & value signal; the enrichment lets you slice it both ways. Add a question next season and it applies retroactively to this one.
Cost & value: the derived lens that unifies the primary categories. Pull every price, fee, and worth-the-money mention across concessions, parking, and merchandise into one derived theme and it becomes the single sharpest signal in the study — 84.0% bottom-two-box, a 2.16 average rating, a 90.8% issue rate on 3,576 comments. It out-scores every survey category for dissatisfaction. That is why value belongs in the interpretation of the primary categories even though it is not a survey question: it is the same fan and the same dollar, seen through a lens the survey does not have.
Bottom-two-box rate: the derived cost & value theme against the seven structured-survey categories. Cost & value is highlighted because it is not a survey question — it is derived from the same text.
The red bar is the derived theme; the gray bars are the survey categories from the front half of this report. Value dissatisfaction runs nearly double the worst survey category — the signal exists in the text whether or not a question asks for it.
bottom-two-box · cost_value (derived) vs primary category_overview · 2024-25 · linked bases vary by category
It amplifies the same primary findings. The value lens does not compete with the survey categories — it concentrates them. In the win/loss layer, value complaints average 8.28 NPS, below even the loss-night average (8.48) and well under result mentions (9.25) — a sharper version of Section 08. Across segments, repeat buyers hit 92.7% value bottom-two-box versus 79.4% for single-game buyers — a wider gap than the food split in Section 09. And at the club level, Dallas pairs a 91.2% value bottom-two-box with a 1,911-comment base, the clearest team value signal in the data.
Derived cost & value bottom-two-box by club, with linked base beside each bar. Dallas is highlighted because its rate holds on far more comments than any other club.
Thin-base clubs can look just as weak, but their rates rest on fewer than 100 comments each. Dallas is the consequential case: an extreme rate across nearly 2,000 comments.
cost_value (derived) bottom-two-box by team · minimum 50 linked comments · 2024-25 · descriptive comparison only
Make things a little less expensive. It’s extremely hard to take my family to a game
And $11 for a soda? That’s obscenely greedy.
| Derived category (from open text) | Base | Avg rating | Bottom-2-box | Research question it answers |
|---|---|---|---|---|
| Cost & value | 3,576 | 2.16 | 84.0% | Is the night worth the spend? |
| Ticketing process | 572 | 3.19 | 40.3% | Is buying and entry a friction point? |
| Game presentation | 971 | 3.40 | 38.2% | Is the show itself landing? |
| Premium | 1,115 | 3.88 | 19.0% | Are premium buyers well served? |
| Venue comfort | 4,651 | 4.05 | 23.8% | Are seats, cleanliness, temperature okay? |
Atmosphere and team-performance categories were also derived from the text but did not join cleanly to the 2024-25 structured spine in this run, so they are held out rather than forced into the comparison.
Seven survey questions produced fourteen enriched categories, and each new lens applies retroactively to the same corpus at no additional fielding cost. Next season’s question set can be different again — a new theme, a new segment cut, a new failure mode — run against feedback fans have already given.
That is the flexibility of deriving categories from raw text rather than hard-coding them into the instrument: the survey defines what you ask, but the enrichment defines what you can learn — and the second set can keep growing long after the season ends.
Methodology & traceability
The primary layer is the seven structured-survey experience questions — entering & exiting, food & beverage, in-game entertainment, merchandise, staff, transportation, venue technology — plus OSAT and NPS as the outcome measures. This report leads with those, per the analysis plan, and defines drivers as the aspect-level subcategories inside them.
The derived layer — cost & value, venue comfort, game presentation, ticketing, premium, and others — was produced by AI enrichment of the open-text answers (“how can we improve,” “what went well”). These add explanatory power but are not survey questions, so they are presented as lenses in the Derived categories section rather than in the primary spine.
The season contains 117,975 structured responses. The win/loss context layer covers 43,732 responses and 40,111 linked open-ended comments. Open-ended feedback is a priority signal from fans who chose to write comments — not an incidence rate for all attendees. Category-level linked bases vary materially, so each category carries its own base.
Bottom-two-box is the standard dissatisfaction lens (predicted category ratings of 1 or 2). Predicted category ratings are text-derived and kept distinct from the structured satisfaction items.
sat_fb sat_transportation sat_venue_tech sat_staff sat_merch sat_in_game_experience sat_entering_exiting_arena nps OSAT buyer_type outcome food_beverage_rating transportation_rating venue_technology_rating staff_rating cost_value_rating wl_driver_food_beverage wl_driver_transportation wl_result_mention