Allergen training rollout for QSR

## The training surface A real allergen training covers four things: (1) the regulated list (EU 14: gluten/cereals, crustaceans, eggs, fish, peanuts, soybeans, milk, nuts, celery, mustard, sesame, sulphur dioxide, lupin, molluscs; US Big 9: milk, eggs, fish, shellfish, tree nuts, peanuts, wheat, soybeans, sesame), (2) the menu-by-allergen matrix (which menu items contain which allergens, including hidden sources like marinades), (3) cross-contact prevention (separate prep surfaces, glove changes, fryer-oil considerations), and (4) the customer interaction protocol (what to say, what to escalate, when to recommend the guest order something else). Most chains cover surface 1 well, surface 2 partially, and surfaces 3 and 4 inconsistently. The customer interaction is where things go wrong publicly. ## Scenario-based is the only approach that works An allergen training that lists the 14 allergens and asks 'name three' produces certificates but doesn't change behavior. The training that does change behavior is scenario-based: the frontliner reads a 3-sentence scenario and picks an action. 'A guest with a peanut allergy is asking about your chicken nuggets. The chicken is fried in shared oil with onion rings that contain wheat but not peanuts. What do you tell them?' The right answer involves the cross-contact risk and the manager-escalation protocol. The wrong answer involves a fast 'they're fine' response. A scenario course covers 15–20 of these patterns and the frontliner internalizes the right reasoning. Aristotl's Socratic-question format was designed for exactly this kind of branching content. The scenario, the answer, the why — all in a 4–6 minute mobile module. ## The matrix problem Every menu item contains a specific set of allergens. When the menu changes (and QSR menus change quarterly), the matrix changes. The training has to update with the matrix. The traditional approach: someone manually updates a laminated chart, photocopies it, ships it to every store. The result is inconsistent versions in circulation and frontliners unsure which chart is current. The fix: the menu-allergen matrix lives as a live document, the training references the matrix, and when the matrix updates the training updates automatically. Aristotl's content engine supports exactly this — the source allergen-matrix doc gets versioned, the courses regenerate when it changes, and HQ tracks which version every employee was trained on. ## The cross-contact training Cross-contact prevention is operational training that can't be fully covered in a course. It needs hands-on reinforcement: separate prep boards for shellfish prep, glove changes between handling allergens, fryer-segregation rules. The course covers the why and the protocol; the manager covers the how, on shift. Both are needed. The operational discipline: every new hire completes the cross-contact course before their first food-handling shift, then has a one-on-one with the manager covering the location-specific protocols (where the dedicated prep surfaces are, how the fryer-oil rotation works at this store). ## Audit and incident records Allergen-related incidents — a customer reports a reaction, an inspector asks for records, a near-miss is reported — all require records. The records need to show: who was trained, on which version of the protocol, when. With a real training platform, this is a 30-second report. With a Sheets file and emails, this is a multi-hour reconstruction. The second most-common audit moment we see in QSR is an allergen one. The first is food-safety. Both reward the operator who has clean digital records and punish the operator with paper or spreadsheets. ## What good looks like A well-run allergen training program has every food handler trained within 5 days of hire, has scenario-based reinforcement that updates with menu changes, has cross-contact protocols reinforced through manager-led shift training, and has audit records exportable in under 2 minutes per location. The operational ROI is in the incidents that don't happen — and that you can prove didn't happen.

Ready to put this into practice?