> ## Documentation Index
> Fetch the complete documentation index at: https://docs.everybite.com/llms.txt
> Use this file to discover all available pages before exploring further.
# The Food Connection
> How guest, food, ordering, and loyalty data connect to create personalized dining
## Four Worlds, One Language
Restaurant data lives in silos. Guest profiles sit in loyalty systems. Menus live in ordering platforms. Nutrition facts hide in spreadsheets. Transaction history scatters across POS systems.
EveryBite connects these worlds—and the key that unlocks everything is **food**.
Who they are, what they want, how they behave
Menus, recipes, ingredients, nutrition, allergens
What they ordered, how they customized, when & where
Points, tiers, tags, segments, campaigns
**The insight:** You don't truly know a guest until you understand their relationship with food. Loyalty points tell you they're valuable. Order history tells you they visit Tuesdays. But *food* tells you they're managing a wheat allergy, building muscle, feeding a family, or celebrating a birthday with something indulgent.
***
## The Four Domains
### 1. Guest
**Who is this person?**
The guest domain captures identity, preferences, and behavior:
| Data | Source | Example |
| ---------------- | ---------------------------------------- | -------------------------------------- |
| **Identity** | Loyalty ID, Passport, device fingerprint | "This is Sarah, a returning customer" |
| **Demographics** | Inferred from behavior patterns | "Young professional, health-conscious" |
| **Preferences** | Explicit settings or observed patterns | "Avoids dairy, prefers high-protein" |
| **Behavior** | Session analytics, interaction patterns | "Browses thoroughly, customizes often" |
But identity alone is shallow. Knowing Sarah is a "Gold member who visits twice a week" doesn't tell you *what she wants to eat*.
### 2. Food
**What can they eat?**
The food domain is the [Ingredient Intelligence](/docs/concepts/ingredient-intelligence) layer—a seven-level hierarchy from menu down to ingredient specification:
| Level | What It Captures |
| ---------------------------- | ---------------------------------------- |
| **Menu** | What's available today |
| **Dish** | What guests see and order |
| **Recipe** | How the dish is assembled |
| **Prep Recipe** | House-made components |
| **Ingredient** | Individual items with allergens |
| **Ingredient Data** | Nutrition facts, dietary classifications |
| **Ingredient Specification** | Exact products and brands |
This depth is what makes EveryBite different. We don't just know "Caesar Salad"—we know the dressing contains anchovy and egg, the croutons have wheat, and the parmesan is dairy. We trace allergens through every layer.
### 3. Ordering
**How do they buy?**
The ordering domain captures transactions and customization:
| Data | Source | Example |
| ------------------ | ----------------------- | ------------------------------------------ |
| **Order History** | POS, ordering platforms | "24 orders, \$18.50 average" |
| **Frequent Items** | Transaction analysis | "Honey Garlic Stir-Fry, ordered 8 times" |
| **Customizations** | Modifier selections | "Always adds extra protein, removes sauce" |
| **Timing** | Order timestamps | "Tuesday lunch, Thursday dinner" |
Order data shows *what happened*. But without food intelligence, you can't answer *why* they customized, or what else they might like.
### 4. Loyalty
**What do their programs know?**
The loyalty domain pulls data from third-party systems:
| Data | Source | Example |
| ----------------- | ---------------------------- | ----------------------------------- |
| **Tier & Points** | Thanx, Punchh, Paytronix | "Gold tier, 2,450 points" |
| **Tags** | Loyalty program segmentation | "birthday\_month", "protein\_lover" |
| **Offers** | Campaign targeting | "Eligible for free dessert promo" |
| **Preferences** | Stated in loyalty profile | "Prefers no onions" |
Loyalty platforms know value and engagement—but they don't understand *food*. They can't tell you if a "high-value customer" can actually eat your new menu item.
***
## Food: The Connective Tissue
Here's the key insight: **Food is how you understand people.**
Without food intelligence, the other domains are disconnected:
| Domain | Without Food | With Food |
| ------------ | ------------------------- | ---------------------------------------------------------------------------- |
| **Guest** | "Sarah visits Tuesdays" | "Sarah avoids wheat, loves high-protein, customizes for her macros" |
| **Ordering** | "She orders the stir-fry" | "She orders the stir-fry because it's her highest-protein wheat-free option" |
| **Loyalty** | "Gold tier, 2,450 points" | "High-value guest who needs wheat-free options—show her the new grain bowl" |
Food turns data into understanding. It answers:
* **Why** does she order that dish? *(It fits her dietary needs)*
* **What else** would she like? *(Other high-protein, wheat-free options)*
* **How** should we personalize? *(Highlight compatible dishes, warn about allergens)*
* **When** is she at risk? *(If we remove her favorite dish, she might churn)*
**"EveryBite"** — The name isn't accidental. Every bite tells you something about the person. What they choose, what they avoid, how they customize—these are signals that reveal who they are and what they want.
***
## How the Domains Connect
### Guest + Food = Personalization
When a guest opens the menu, we know their dietary restrictions before they search:
```
Guest: "wheat-free", "high-protein"
↓
Food: Traverse ingredient hierarchy
↓
Result: Dishes safe for this guest, ranked by preference match
```
The guest sees a personalized menu from the first screen—not because we guessed, but because food intelligence lets us trace compatibility through every ingredient.
### Ordering + Food = Understanding
When a guest customizes, we understand *why*:
```
Order: "Remove croutons, add extra chicken"
↓
Food: Croutons contain wheat, chicken adds protein
↓
Insight: Guest is likely wheat-free and protein-focused
```
This inference feeds back into the guest profile, improving future personalization.
### Loyalty + Food = Actionable Segments
Loyalty tags become meaningful when combined with food:
```
Loyalty: "Gold tier", "weekday_regular"
↓
Food: Observes wheat-free, high-protein ordering pattern
↓
Segment: "High-value health-conscious regular"
↓
Action: Send new protein bowl promo, not pasta special
```
Without food, you'd send the pasta promo to someone who can't eat it.
### Guest + Ordering + Loyalty + Food = GuestIQ
When all four domains connect, you get complete guest intelligence:
**Guest** (identity & behavior) + **Ordering** (history & customization) + **Loyalty** (tags & tiers)
The **Ingredient Intelligence** layer analyzes everything through the lens of food—what they can eat, what they prefer, what they avoid.
**Complete Profile** · **Smart Segments** · **Targeted Dishes**
This is what [GuestIQ](/api/guestiq/overview) returns at session creation—the guest's complete profile, understood through the lens of food.
***
## What This Means in Practice
### For Menu Personalization
A guest opens your app. Before they search, you already know:
* **Their restrictions:** Wheat-free, peanut allergy
* **Their preferences:** High-protein, prefers lunch portions
* **Their patterns:** Usually customizes, price-sensitive
* **Their favorites:** Ordered the Power Bowl 6 times
The menu they see is already filtered, with compatible dishes highlighted and allergen warnings in place.
### For Campaign Targeting
You're launching a new high-protein grain bowl. Who should see the promo?
| Traditional Targeting | Food-Informed Targeting |
| ------------------------ | ----------------------------------------------- |
| "Active loyalty members" | "Wheat-free + high-protein segment" |
| "Frequent visitors" | "Guests who order protein modifications" |
| "High spenders" | "Guests whose favorites are being discontinued" |
Food intelligence makes campaigns precise, not spray-and-pray.
### For Staff Alerts
A VIP guest walks in. Your host stand shows:
> **Sarah M.** — Gold member, 24 visits
>
> * Wheat allergy (verified)
> * Prefers: Power Bowl, Stir-Fry
> * Usually adds: extra protein
> * Note: Birthday this month
The staff can greet her by name and recommend something she can actually eat.
### For Product Development
Which dishes should you add to the menu?
| Question | Food Data Answers |
| ------------------------- | ---------------------------------------------------------- |
| "What's missing?" | "12% of guests are wheat-free but only 3 dishes fit" |
| "What's underperforming?" | "This dish matches many profiles but has low reorder rate" |
| "What's trending?" | "High-protein customizations up 40% this quarter" |
***
## The Data Model Summary
| Domain | What It Knows | Key Insight |
| ------------ | --------------------------------- | ----------------------- |
| **Guest** | Identity, demographics, behavior | *Who* the person is |
| **Food** | Ingredients, nutrition, allergens | *What* they can eat |
| **Ordering** | Transactions, customizations | *What* they chose |
| **Loyalty** | Points, tiers, tags | *How valuable* they are |
**Food is the key.** It transforms:
* Anonymous visitors into understood guests
* Transaction data into dietary insights
* Loyalty tiers into actionable segments
* Generic menus into personalized experiences
***
## Learn More
Deep dive into the seven-layer food hierarchy
How we deliver complete guest profiles at session creation
Personalized menu search powered by food intelligence
Start sessions and receive guest intelligence