Brand Standards
How to score subjective criteria (like hospitality) objectively
Turn vague standards like friendly service and inviting atmosphere into observable, scorable behaviors that consultants grade the same way at every store.
Every franchise brand manual has a section on hospitality. It says guests should feel welcome, team members should be friendly and the atmosphere should be inviting. Then the audit checklist asks the consultant to score hospitality from 1 to 5, and two consultants visiting the same store on the same day give it a 2 and a 4. The franchisee who got the 2 is right to feel the score is arbitrary.
Hospitality is not unmeasurable. It is just usually written in words that describe a feeling rather than a behavior. The fix is to translate each feeling into things an auditor can see or hear, and score those.
Replace adjectives with behaviors
Take every subjective line in your rubric and ask: what would I have to see or hear to know this standard was met? If two trained people cannot agree on the answer, the line is not ready to score.
| Subjective standard | Observable behaviors |
|---|---|
| Guests are greeted warmly | Greeting within 5 seconds of entering or reaching the counter; eye contact; greeting includes the brand welcome phrase or equivalent |
| Team is friendly and engaged | Team member smiles during the order; uses the guest's name if it is on the order; thanks the guest at handoff |
| Order taking is helpful | Repeats the order back; offers the current featured item once; answers a menu question accurately |
| Atmosphere is inviting | Music at brand volume level; lighting all working; no team member personal items in guest view; no team conversations audible in the dining room |
| Guest leaves happy | Order accurate; any issue resolved on the spot by a team member without the guest needing to ask twice |
Each behavior is binary or close to it. The greeting happened within five seconds or it did not. The order was repeated back or it was not. Consultants can still disagree at the edges, but the range of disagreement shrinks from three points on a five-point scale to the occasional borderline case.
Choose the right scale
Where you cannot reduce a standard to yes or no, use a short anchored scale, where every point has a written definition and ideally a reference example. Avoid five-point scales for anything you want consistency on; consultants cluster on 3 and 4 and the scale loses meaning.
- Pass or fail: for behaviors that either happened or did not, such as a greeting or a repeated order.
- Three-point anchored: for quality judgments, with each point defined. For example, 0 means no greeting, 1 means a greeting that is late or scripted without eye contact, 2 means a greeting within five seconds with eye contact.
- Count based: for things that can be tallied, such as the number of tables uncleared for more than five minutes during the walk.
Sample more than one interaction
A single guest interaction is a weak basis for a hospitality score. The consultant might have caught the best team member on their best moment or the newest hire on their first shift. Where the visit allows, observe three interactions at the counter or drive-thru window from a position where you can hear, and score the share that met each behavior. Three of three greetings within five seconds is a pass. One of three is a coaching conversation, not a guess.
Separate hospitality from the rest of the score
Many brands fold hospitality into a single total with food safety and facility condition. That hides problems in both directions. A store with spotless operations can mask a cold, transactional team, and a warm team can mask a failing walk-in cooler. Report hospitality as its own zone score so it can be coached on its own terms. In Gazi Wellness, each zone is weighted and reported separately, and the overall score is a roll-up rather than a single number entered by the consultant.
Calibrate with real examples
Behavioral definitions get you most of the way. Calibration gets you the rest. Once a quarter, have your consultants score the same set of recorded or described interactions independently, then compare. Discuss every item where scores differ and update the written definitions with the clarification. Over a few cycles the definitions become specific enough that new consultants can score consistently in their first month.
- Pick five to eight interactions that represent common borderline cases.
- Have every consultant score them independently against the current rubric.
- Compare scores and discuss each item with disagreement.
- Update the definition or add a reference example that resolves the disagreement.
- Track the agreement rate over time, aiming for at least 85% agreement on each line.
What can and cannot be scored from photos
Photos help with the environmental parts of hospitality: lighting, cleanliness of the dining area, condiment station setup, uniform and name tag compliance. Computer vision can flag a burned-out fixture or an overflowing trash receptacle reliably. It cannot judge whether a greeting was warm. Be clear in your rubric about which items are photo verified and which are observed, and do not ask a tool to score what only a person can see. Gazi Wellness's Vision AI scoring is applied only to items that are configured as photo-verifiable; observed behaviors stay with the consultant.
Coach to the behavior, not the number
The real benefit of behavioral scoring shows up in the coaching conversation. Telling a GM that hospitality scored 2 out of 5 gives them nothing to act on. Telling them that two of three guests were not greeted within five seconds, and that the drive-thru window team did not repeat orders back, gives them two specific things to train at the next pre-shift huddle. Franchisees accept a score they can see the reasons for, and they can fix what they can name.