Glossary

(positive mentions − negative mentions) / (positive + negative + mixed mentions), on a −1 to +1 scale.

An aspect score is the per-dimension sentiment net for an entity. We take every customer review that mentions the dimension, classify each mention as positive, negative, or mixed via the LLM extractor, then compute (pos − neg) / total. A score of +0.50 means customer mentions of that dimension are net-positive at a 3:1 ratio; −0.30 means net-negative at roughly 2:1; 0.00 means evenly split.

Aspect scores carry more signal than the raw 5-star Google rating because they isolate what specifically went right or wrong. A 4.2★ chain with an aspect score of +0.90 on staff but −0.40 on equipment has a different operational profile than another 4.2★ chain with the inverse pattern, and the strategic implications for an operator or PE buyer are different.

See also: wash quality · wait time · staff · equipment reliability · facility cleanliness · membership value · per wash price

Aspect score for whether the wash itself delivered a clean car.

Captures customer commentary on the actual cleaning result: spots left behind, streaks, dirt on wheels, condition of the interior after a full-service. Aggregated as an aspect score on the −1…+1 scale.

Wash quality is the most-cited aspect across the WashIndex corpus and the strongest single predictor of repeat visit intent. Operators investing in better wash chemistry, brush quality, drying systems, or staffing for detail shops typically see this aspect move first.

See also: aspect score

Aspect score for line length, throughput, and the time it took the customer to get through the wash.

Customer commentary on how long they waited, whether the line moved efficiently, and whether throughput felt acceptable. Express tunnels typically score higher on wait time than full-service operations because the format is built for throughput.

Wait-time scores correlate strongly with weekend volume per bay/tunnel and with format. A wait-time score that drifts negative over time is often the earliest signal of either capacity strain or operational degradation (equipment issues slowing throughput).

See also: aspect score

Aspect score for staff friendliness, helpfulness, and professionalism.

Captures customer commentary on attendants, cashiers, detailers, and any staff member the customer interacted with. Full-service and detail-shop operators typically have higher staff scores because staff interaction is built into the experience; express tunnels often have lower staff visibility but can still score well if the prep crew is responsive.

Staff is the highest-rated aspect across the corpus on average. Operators with notably low staff scores often have systematic training or compensation issues rather than individual personnel problems.

See also: aspect score

Aspect score for whether vacuums, dryers, conveyors, and other equipment worked.

Customer commentary on broken vacuums, missing or broken dryers, conveyor issues, malfunctioning tire shine machines, and similar equipment problems. This is one of the few aspects that frequently scores negative across the corpus — broken equipment is a common complaint pattern.

Equipment scores correlate with chain age, deferred maintenance, and operator philosophy. A chain with damage-cluster signal (high reviewer-reported damage rate) almost always has weak equipment-reliability scores as well; they're related manifestations of the same operating deficit.

See also: aspect score · damage rate

Aspect score for the appearance and upkeep of the physical facility outside of the wash itself.

Captures customer commentary on litter, paint, signage condition, restroom cleanliness, lot maintenance, and overall property appearance. Distinct from wash quality (which is about the cleaning result on the car).

Facility cleanliness usually correlates with operator capex discipline and management cadence. Multi-site operators with sharp facility-cleanliness scores at older sites typically have a documented refresh/repaint cadence built into their operating model.

See also: aspect score

Aspect score for how customers feel about the unlimited-membership program economics.

Customer commentary on whether the membership felt worth the monthly price. Includes pricing critiques, value comparisons to competitors, and complaints about hidden fees or unclear tier structure.

Distinct from membership cancellation friction (which captures pain when customers try to leave). A chain can have high membership-value scores while still having elevated cancellation friction, or vice versa.

See also: aspect score · membership cancellation friction

Aspect score for how customers feel about non-membership single-wash pricing.

Customer commentary on the price of a single wash for non-members. Captures both absolute price perception and relative-to-quality assessments. Different formats have different price expectations: an express tunnel at $15/wash and a detail shop at $80 can both score well on price because customers calibrate to format.

Negative per-wash-price scores at the chain level usually indicate either pricing that's outpaced format expectations or a recent price hike that customers haven't accepted. Sharp negative shifts often correlate with rating drift episodes.

See also: aspect score · rating drift

The fraction of LLM-extracted reviews that mention vehicle damage at the location, chain, or market.

Computed as (count of reviews with a damage mention) / (total LLM-extracted reviews) for the entity. Damage mentions are detected as a discrete event field by the extraction pipeline, not inferred from sentiment.

Across the WashIndex chain corpus, the median chain-category damage rate sits in roughly the 1–3% range. Chains running 1.5× to 5× above median (3–10%+) are typically termed damage-cluster operators and surface as quality outliers on chain detail pages.

Damage rate is the most-cited single signal by P&C insurance underwriters and PE diligence teams because it has direct dollar implications (claims cost) and is hard to mask through marketing.

See also: equipment reliability · scored reviews

Among reviews that discuss the unlimited-wash membership program, the fraction that report difficulty canceling or getting refunded.

Computed as (count of mem-discussing reviews with cancellation-friction signal) / (count of mem-discussing reviews). The denominator restricts to reviews that mention membership at all — otherwise the rate would be diluted by drive-up customers who never engage with the program.

Cancellation friction is a leading indicator of churn risk and brand-trust erosion. It's distinct from membership-value sentiment: a chain can be priced fairly (good value) but still make canceling difficult (high friction), or vice versa.

Many chains' cancellation-friction rate cluster between 8% and 18% of mem-discussing reviews. Rates above 25% indicate either policy issues (in-person-only cancellation, mandatory call queues, refund delays) or systematic process problems that surface across multiple sites.

See also: membership value · membership mention rate

Fraction of LLM-extracted reviews that complain about pushy upselling at the wash entrance or by staff.

Computed as (count of reviews mentioning upsell-pressure complaints) / (total LLM-extracted reviews). Captures the customer-friction dimension of operator monetization strategies. Express tunnels with attendants who push upgrades, or full-service operators who recommend add-ons, can drive elevated upsell-pressure rates.

Upsell pressure correlates with revenue-per-wash but also with rating drift. Operators that move membership-mention rate up while moving upsell-pressure rate up have typically over-rotated on aggressive sales tactics that hurt the customer experience.

See also: membership mention rate

Fraction of LLM-extracted reviews that mention the unlimited-membership program at all.

Computed as (count of reviews mentioning the membership) / (total LLM-extracted reviews). Captures how visible and salient the program is to customers — a leading indicator of membership penetration in the customer base.

Membership mention rates vary by format. Express tunnels selling unlimited monthly subscriptions typically run 8–15% mention rates; full-service operators with no recurring program often run 1–3%.

Used as the denominator for membership-cancellation-friction and similar mem-conditional metrics so that chains without significant membership programs aren't distorted.

See also: membership cancellation friction

Per-entity star rating computed as the review-weighted mean across all locations in the entity.

For a chain or market with N locations, the weighted rating is Σ(locationRating × locationReviews) / Σ(locationReviews). This differs from a naive mean of location ratings because high-review-count locations have more statistical weight — which is the right behavior, since they carry more customer signal.

Compared against Google's per-listing star rating, the weighted average controls for the fact that small-volume sites can drift to extreme values. A chain with one 5.0★/3-review site and one 4.0★/3000-review site has a weighted average near 4.0★ (the right answer) rather than 4.5★ (a naive mean).

See also: scored reviews

The subset of total Google reviews that have been processed through WashIndex's 55-field LLM extraction pipeline.

WashIndex tracks two review counts per entity: total reviews (what Google displays publicly) and scored reviews (the subset that has passed through structured extraction). Aspect scores, damage rates, and other event-detection metrics are computed from the scored-reviews denominator, not the total-reviews count.

The extraction pipeline excludes spam-flagged reviews and reviews lacking enough text signal to extract from. For chains with high scored-review counts (10K+), the resulting metrics are statistically reliable; for chains with sparse scored-review counts (under 100), confidence intervals widen.

See also: spam flagged reviews

Fraction of mem-discussing reviews where the reviewer explicitly identifies as an unlimited-wash member.

Captures the share of the customer base that has converted to the recurring monthly program, inferred from review-text self-identification. A leading indicator of customer LTV — unlimited members visit far more often than per-wash customers.

Membership penetration is the metric most commonly looked at by PE buyers evaluating chain economics; a chain with 35%+ unlimited-membership penetration has a substantially different revenue profile than one running 10%, even at similar per-wash pricing.

See also: membership mention rate

Conveyor-driven tunnel format: the customer stays in the car while it's pulled through automatic equipment.

Throughput-optimized format with conveyor systems, automated cleaning equipment, and (usually) attended pay stations or membership scanners at entry. Wash takes 3–5 minutes. Express tunnels are the format most associated with recurring monthly memberships and the format that has driven most US car-wash chain rollup over the past decade.

Express tunnels typically have higher wait-time aspect scores and lower staff aspect scores than full-service operators (less staff interaction by design).

See also: full service · in bay automatic

Staffed format that combines automated wash equipment with hand-finishing: vacuuming, drying, and interior wipe-down by attendants.

Hybrid format: the wash itself runs through automated equipment, but staff finish the car after — vacuum interior, dry exterior, dress tires, wipe down dash. Wash takes 15–30 minutes. Full-service operators typically run higher per-wash prices than express tunnels because the staff intensity is higher.

Full-service is most common in legacy markets (Northeast, parts of California) where it predates the express-tunnel rollup wave. Customer expectations and aspect-score profiles differ significantly from express — full-service customers care more about staff and result quality, less about throughput.

See also: express tunnel · detail shop

Hand-finished format focused on full interior + exterior cleaning, paint correction, ceramic coating, and similar premium services.

The slowest, most labor-intensive format. Each car gets a multi-hour treatment — typically by appointment — including interior shampoo, paint clay-bar treatment, ceramic coating, or full-service detailing. Per-job pricing runs from $80 to several hundred dollars.

Detail shops typically score highest on wash quality and staff aspect dimensions, and lowest on wait-time (customers wait hours for the result). Different customer occasion than tunnel formats — detail shops compete on result quality, not throughput.

See also: full service

Coin-operated or app-paid format where the customer operates the wand and brushes themselves.

Drive-in bay with a high-pressure wand, brush, foam, and rinse functions controlled by the customer. Pay-as-you-go (typically $2–5 per cycle). Common in rural and small-metro markets where express-tunnel rollouts haven't reached.

Self-serve bays score very differently on the aspect framework — low staff (no staff to score) but often surprisingly high facility-cleanliness if maintained well, since customers can see the bay condition before pulling in.

See also: in bay automatic

Single-bay format with automated equipment that moves around a stationary car.

The car drives in and parks; the equipment (usually a touchless gantry or rolling brush carriage) moves around it. Wash takes 3–6 minutes. Cheaper to build than a tunnel and often co-located at gas stations or convenience stores.

Aspect-score profile sits between express tunnel (no human interaction) and self-serve bay (driver stays in car the entire time). Wash quality typically scores below tunnel formats.

See also: express tunnel · self serve bay

Format where staff hand-wash the entire car without going through automated tunnel equipment.

Less common in the US than the formats above; more prevalent in coastal urban markets where labor is available and customers pay a premium for hand-touch service. Wash takes 30+ minutes. Per-wash pricing runs $25–60.

Customer expectations are closer to detail shop than express tunnel; the format competes on quality and care rather than throughput.

Operator that comes to the customer's location with portable equipment.

Truck-based or van-based detailing service that travels to residential or commercial addresses. Most often a sole-proprietor or small-team operation. Pricing varies widely; aspect-score profiles are distinct because the location convenience is built into the value proposition.

Format specialized for commercial trucks, RVs, and fleet vehicles.

Operators with equipment scaled to handle vehicles too large for a standard tunnel. Bays or drive-through tunnels are sized for 18-wheelers, motorhomes, or commercial trucks. Different customer base (commercial / fleet) and different unit economics than passenger car formats.

Core-Based Statistical Area / Metropolitan Statistical Area. Federally-defined market boundaries used by Census, OMB, and the WashIndex per-metro analyses.

A CBSA is a county-level geographic delineation centered on an urban core. Metropolitan Statistical Areas (MSAs) are CBSAs with at least one urban area of 50,000+ population. WashIndex profiles all 392 US MSAs per the 2023 OMB Bulletin 23-01 delineation.

Each MSA can span multiple states (NYC = NY+NJ+PA; Cincinnati = OH+KY+IN). Inclusion in a metro is determined by point-in-polygon test against the official US Census TIGER/Line county boundaries.

See also: top10 chain share

Fraction of the metro's car wash locations operated by the top 10 chains, by location count.

A market-concentration metric. Top-10 share around 0.20 indicates a fragmented market dominated by independents; around 0.50+ indicates a chain-consolidated market where rollup has progressed significantly. The US-metro median typically sits around 0.30.

WashIndex uses top-10 share (alongside independent share) to detect the 'consolidated' archetype in its MSA narrative engine.

See also: independent share · format dominance

Fraction of the metro's car washes operated as independents or single-location operators (not part of any tracked chain).

The complement of chain share — but not exactly 1 minus the top-10 share, because smaller chains (top 11–N) also exist. Independent share captures the long-tail layer that hasn't been touched by rollup.

Markets with 0.80+ independent share are termed 'fragmented' in the WashIndex archetype framework; 0.50 or lower indicates significant chain consolidation. Most US markets sit in the 0.65–0.80 range.

See also: top10 chain share

When a single operating format (express tunnel, detail shop, etc.) accounts for 55%+ of formally-classified car washes in a market.

Detected as the share of one format among the format-classified subset (excluding 'unclassified' sites). A market is termed format-dominant when one format runs 55%+; the WashIndex narrative engine routes these pages through a dedicated archetype.

Format-dominant markets are interesting because they tell you something about how the local customer base prefers to wash their car. Detail-shop-dominant cities indicate hand-finish demand; express-tunnel-dominant cities indicate throughput-oriented customers.

See also: express tunnel · detail shop

Count of competing car wash locations within a haversine distance from a target site.

WashIndex computes competitor counts for each tracked location within three concentric rings: within 1 mile, within 3 miles, within 5 miles. Computed by haversine distance against the WashIndex national location index across the location's state plus immediately adjacent states. Sister sites in the same chain are excluded.

Competitive rings provide trade-area density context for site selection, acquisition diligence, and operator-quality comparisons. A site with 0 competitors within 1 mile has very different competitive dynamics than one with 8.

See also: drive time isochrone

Polygon enclosing every location reachable within a given drive time from a starting point.

More accurate than radial distance for trade-area analysis because it accounts for actual road networks, highway geometry, and barrier features (rivers, mountains). The WashIndex platform supports 5-, 10-, and 15-minute isochrones with real road-network routing.

Used on the paid platform; not currently in the per-MSA or per-city HTML reports. The static competitive ring metric (1/3/5 mile) is the public-data version.

See also: competitive ring

A sustained directional change in a chain's monthly average rating, detected by rolling-window comparison.

WashIndex's anomaly detection compares the rolling rating window (most recent 12-14 months) against an earlier baseline window. A drop of 0.15 stars or more, sustained over multiple months, registers as a rating-drift event with an associated start month, duration, magnitude, and dominant complaint theme.

Rating drift events surface as the lead signal on chain pages in the 'drift-detected' archetype. Useful for spotting chains where operating quality has degraded materially since acquisition, leadership change, or roll-up-driven scaling pressure.

See also: weighted avg rating

Reviews identified by the WashIndex pipeline as spam, paid, incentivized, or otherwise non-genuine.

Detected by a separate spam-classification model that runs on review text + metadata (timestamps, reviewer profile signals, review length distributions). Spam-flagged reviews are excluded from aspect scores, damage rates, and all other aggregates — but the count is preserved and disclosed on the underlying chain JSON.

Most chains have spam-flagged review counts in the low single-digit-percentage range. Outliers with 10%+ spam-flag rates often correlate with chain-driven incentive campaigns (asking customers to leave 5★ reviews in exchange for discounts), which can be detected by the model.

See also: scored reviews

Reviews processed through WashIndex's 55-field structured-extraction pipeline.

Every customer review WashIndex ingests is run through an LLM with JSON-schema enforcement that extracts 55 distinct structured fields — damage event yes/no, wait-time complaint yes/no, staff sentiment positive/negative/mixed, and so on. This is the input to all aspect scores and event-rate metrics.

The 55-field schema was designed for institutional decision use cases (PE diligence, insurance underwriting, operator benchmarking). Fields that don't help one of those audiences don't earn a slot.

See also: scored reviews · aspect score

Composite score combining population density, household income, home values, traffic counts, competition, and ratings into a single site-selection metric.

Used on the WashIndex platform's site-selection workflow. Not exposed in the public MSA / city HTML reports — those provide the underlying inputs (population, income, competitive ring counts) but not the composite.