Bandana’s rental station rating is an assessment of public customer feedback, on a scale from 1 to 5. It is not the same as a station’s Google rating, an inspection certificate or a guarantee of your rental experience. A station receives a rating only when usable evidence is available.
What we assess
We look for feedback about the exact depot, using the supplier’s name, city, country and known local or alternative names. Available evidence can include Google reviews and other public review sites, travel forums and blogs. Not every station has evidence from every source.
Each review contributes only to the topics it actually discusses. The ten satisfaction topics have these relative weights, totalling 95 points. These are weighting points, not percentages of every final rating:
- RV quality & model years: 12 points
- Equipment & amenities: 8 points
- Maintenance: 12 points
- Cleanliness: 8 points
- Supplementary gear (bedding, kitchen, chairs): 5 points
- Pickup / handover: 10 points
- Service during the trip: 12 points
- Return process: 8 points
- Fair treatment: 12 points
- Advertised promises kept: 8 points
How evidence becomes a rating
AI helps identify topics and assess the sentiment of sourced feedback. A fixed calculation combines that evidence; AI does not directly choose the final station rating.
- Missing information is not a zero. A topic with no usable evidence is marked “Source not found” and excluded from the topic average.
- Station and fleet evidence are distinguished. Where both exist, a small station sample leans more on evidence about its fleet brand. Fleet-only topics have partial support. If only network evidence is available, the rating has low confidence.
- Network influence is limited. When station or fleet evidence is available, the wider network can adjust the combined score by no more than 0.30 points in either direction.
- Recent evidence carries more weight. Reviews up to 12 months old receive full age weight; 13–24 months receive 90%, 25–36 months 75%, 3–5 years 50%, 5–8 years 25%, and older reviews 10%. Unknown dates receive 50%.
- Evidence quality matters. Detailed, dated feedback clearly tied to the depot receives more weight than vague feedback. Material classified as unusable, such as spam, duplicates or feedback about another location, is excluded.
- Overall Google ratings also contribute. A station’s own listing average is blended with topic evidence. Its review count is capped at 100 for this calculation, at a quarter of the weight per review. This helps balance detailed complaints against the broader rating sample without allowing review volume unlimited influence.
Confidence is separate from satisfaction
Confidence runs from 0 to 100 and reflects the weighted amount of station and fleet evidence, how many topics have evidence, and known years of operation. Operating history contributes to confidence only—not to satisfaction. Network-only confidence is capped at 25.
Confidence is an indicator, not a statistical probability. Several topic findings can come from one customer review, and many findings can come from one source. A high confidence score does not guarantee broad source diversity or eliminate search bias.
How to use the result
Read the rating together with its summary, confidence and available sources. Public feedback may be incomplete, outdated or unrepresentative, and summaries reflect the evidence available when assessed. Not all stations have been assessed, and refresh dates can vary. Always check the supplier’s current terms and vehicle details before booking.
Disclaimer: All the information is based on public information on the web. All imagery sourced under Creative Commons licenses with attribution provided. Always check trail conditions and opening hours before travel.
