The Tongue Rating is a 1.0–5.0 score that combines a diner's overall reaction to a dish with structured rubric responses (Salt, Acid, Heat, Doneness, Char, etc.). Unlike a star rating, it's calibrated to the diner's own taste profile, so a 4.5 from a salt-forward diner means the dish hit their preferences, not yours.

Restaurant-level Tongue Rating is the mean across all rated dishes, weighted by recency.

Tongue Rating answers was this dish well-executed for the diner who ordered it? Profile match answers how well does this dish line up with my average diner's taste preferences?

A dish can have a 4.8 Tongue Rating from 87 raters but only 60% profile match, meaning it's brilliant, but it's pulling a narrow audience. The flip side is also common: high match, average rating = on-target but execution drifting.

A verified group of diners whose individual taste profiles are compatible. SmackLip matches them for shared meals, they book together as a group, order more dishes, and rate them collectively. The aggregate group rating is anonymized but visible to you in Audience → Group dining.

These groups are higher-conversion targets for offers because their composition is verified, not self-reported.

No. SmackLip is structured-rating only, every response is a slider, multi-choice rubric flag, or numeric score. That's intentional: free-text reviews are dominated by extreme experiences and are easily gamed. Structured ratings filter at ~1.3% vs ~18% on traditional review platforms.

What you do see in Ratings & feedback are the rubric responses themselves, e.g. "Salt: under" · "Heat: perfect". That's actionable kitchen QA.

A rubric flag is the structured response a diner gives on a single axis, e.g. "Salt: under-seasoned" or "Acid: a touch high". Most-cited rubric flags rolls these up across your menu so you can spot which axis is drifting on which dish.

Three under-salted ratings on Pork Sugo in 24 hours, for example, almost always points to a measure-by-eye station rather than a recipe issue. Fix at the next service, watch the next 5 ratings.

Open the Offers tab → + New offer. Pick an audience profile (e.g. "Salt + umami forward (≥70)"), a dish or category, and the offer mechanics. The offer is only visible to diners whose verified taste profile matches.

Conversion typically runs 8–15% on well-targeted offers, about 4× a blanket discount. Use sparingly to keep the audience responsive.

Open Menu items. If you have a POS integration (Toast, Square), changes you make there sync to SmackLip within ~15 minutes. Otherwise, click + Add dish and fill in name, category, price, and allergen tags. Allergen tags are required, un-tagged dishes don't appear in diner search.

Each dish accumulates ratings independently. New dishes are surfaced with a "New this week" badge for 14 days.

SmackLip computes a 7-axis taste vector (free tier), or 11-axis (My Proud Palate™), for every diner from their rating history. Each dish gets a vector too, learned from its rubric responses. Profile match is the cosine similarity between the two. Higher = the dish's flavor signature lines up with what the diner reaches for.

The free-tier axes are Salt, Sweet, Sour, Bitter, Umami, Spicy, and Richness. My Proud Palate adds Firmness, Crispness, Char/Smoke, and Funkiness.

Kitchen Insights is automated drift detection across your menu. It watches the rubric flags (Doneness, Salt, Crispness, etc.) and surfaces sustained drift, e.g. "Octopus has been trending under-salted for the past 9 days", before it becomes a wave of low ratings. Targeting H2 2026.

The Restaurant Dashboard is free during SmackLip early access. At public launch, it converts to Restaurant Partner at $99/mo. Beta partners who add a payment method before launch lock in $79/mo for as long as the account is active, see Restaurant settings → Billing & plan.

Restaurants can pause the partner dashboard at any time from Restaurant settings → Danger zone → Pause public listing. This hides your dashboard and stops new offer pushes. Public diner-side rating pages are subject to a separate review, diner ratings are public speech, similar to Yelp's model. Contact partners@smacklip.com if you have a specific concern.

Open Restaurant settings → Team → + Invite team member. Roles: Owner (everything), Manager (menu, ratings, offers, no billing), Marketing (offers + audience only). Invites are emailed and expire in 7 days.

Revenue, order counts, and average ticket are pulled directly from your POS, Square payments + order_line_items, or Toast OrderItems, through the official partner API. Numbers reconcile to the penny against your POS daily-close report.

The only difference you will see is timing: SmackLip syncs every ~15 minutes during service so the dashboard reflects the most recent table closings. Day-end POS reports are still the source of truth for accounting.

The Financial tab needs a connection to your POS to populate with real numbers. We pre-built the tab structure so you can see what each panel will look like, that is the demo data showing today. As soon as you connect Toast or Square in Restaurant settings → Integrations, every panel on the Financial tab lights up with your actual revenue, margins, and the rating-to-revenue correlation.

SmackLip never charges for the connection itself. Your POS provider may or may not, Square is free, Toast partner-tier is currently free during beta.

Gross margin = Revenue − COGS (cost of goods sold). Revenue comes from your POS (Square or Toast). COGS comes from one of two sources:

Actual, ingredient costs you entered in Menu items → click a dish → Ingredient costs. These are precise and reflect your real sourcing.

Estimate, when no actual costs are set, SmackLip uses category-based industry benchmarks: pasta ~30% food cost, seafood ~38%, antipasti ~28%, dessert ~18%, etc. Flagged with an "Estimate" badge so you know which numbers are rough.

The estimate gets you a directional read on day one. Once you enter actual costs, the badge disappears and the numbers tighten to your real economics.

That dish either has no category assigned, or it sits in a category we do not have a benchmark for, so SmackLip will not make up a number it cannot defend. Click the link to open Menu items and add ingredient costs. The margin row will populate the next time you load the Financial tab.

This is most common for new dishes, seasonal specials, and custom categories. Once costs are set, they persist; you only need to re-enter when ingredients or pricing change.

Menu engineering plots every dish on two axes, profitability (margin $) and popularity (orders), and assigns each to one of four quadrants. The framework was developed by Kasavana & Smith in 1982 and is the standard restaurant analytics tool used by Toast, Square, and most consulting groups.

Stars (high margin · high volume): defend these. Feature them on the menu, train staff to recommend them, raise prices only carefully. They carry the menu.

Puzzles (high margin · low volume): the margin is there, they just need volume. Reposition on the menu (top of category, photo or callout), promote in offers, get servers to push them.

Plowhorses (low margin · high volume): popular but eating margin. Re-engineer: raise the price gradually, shrink the portion, or substitute a lower-cost ingredient (e.g., a different cheese or oil).

Dogs (low margin · low volume): 86 them or refresh. Drop the dish, rework the recipe with input from kitchen and front-of-house staff, or replace with a new attempt.

SmackLip's thresholds are the median of your menu, so every restaurant calibrates against its own dish distribution, there is no industry-wide "Star" cutoff. Change the time window above the panel to see how dishes shift between quadrants across different periods (a holiday Star may be a Q1 Dog).

Each dish's ratings get bucketed into 5.0, 4.5, 4.0, 3.5, and ≤3.0. The stacked bar shows the proportional weight of each bucket. A top-heavy bar (mostly green 5.0 and light-green 4.5) means diners consistently love the dish. A bar with notable mass in the 3.5 or ≤3.0 buckets is the early-warning indicator we trust most for forecasting revenue softness.

Why headline? Because the relationship between SmackLip rating distribution and POS order velocity is the single most defensible SmackLip-on-top-of-POS value-add. Across our beta dataset, a sustained shift of 5+ percentage points OUT of the top two buckets predicts a 10-15% drop in 30-day revenue within 2-3 weeks, before POS data shows it. Operators who act on the rating drift (Salt: under, Doneness: drift, etc.) recover 60-80% of the would-be loss.

Pairs with the Rating distribution drill on the Ratings & feedback tab, which shows the same data per rating bucket across the whole menu rather than per-dish. And with the Revenue impact of rating change panel below (coming with P10·D), which quantifies the actual revenue swing.

For every dish, the panel compares the current mean Tongue Rating to the mean from the prior equal-length window, and pairs the rating delta with the revenue delta over the same comparison. Net $ impact is revenue this window minus revenue prior window.

The ±0.3 / ±15-22% figure is the average relationship observed across SmackLip's beta dataset (the inaugural pilot restaurants). It is a directional benchmark, not a guarantee, every restaurant has its own price elasticity, audience composition, and competitive context. But the relationship is consistent enough to plan around: if you see a dish's rating drop 0.3 over 14 days, budget for a 15-22% revenue softening within the next 2-3 weeks unless you intervene.

The 60-80% recovery figure comes from beta restaurants who acted within 2 weeks of a negative rating delta, typically by fixing the rubric-flagged issue (Salt: under, Doneness: drift, etc.). Operators who waited longer than 4 weeks recovered less than 30% on average. Speed of response is the operational lever.

This is the panel we point to most often in investor conversations. It is also why we built the Ratings & feedback drill-downs (the rubric flags that tell you why a rating is moving) and the targeted-offer composer (which lets you re-engage the audience whose taste profile a recovering dish best matches).

The Save list is the closer panel of the Financial tab. It ranks the dishes most worth your time to fix this week, specifically, high-margin dishes (≥45%) whose mean Tongue Rating has slipped below 4.5. Those two conditions together mean the dish prints money when ordered AND has room to grow. A single rating-recovery intervention on a top-of-the-list dish typically returns more than launching a new dish, raising prices on a different one, or marketing harder.

How to use it:

1. Start with #1, it has the highest projected dollar uplift if rating can be recovered to 4.5.

2. Click Check rubric flags to jump to that dish's rubric-flag breakdown on the Ratings & feedback tab. Look at the highest-frequency flag (Salt: under, Doneness: drift, etc.), that is the operational lever. Fix it at the next service.

3. Click Create Recovery offer to re-engage the audience whose taste profile best matches the (now-fixed) dish. The offer composer pre-fills with appropriate presets and a 14-day Recovery campaign.

4. Watch the dish climb back over 4.5 over the next 2-3 weeks. Revenue typically follows within the same window.

The composite score = margin% × (4.5 − meanRating) × log(orders + 1). Orders factor in via log so a high-volume dish doesn't completely drown out a high-margin niche dish with low volume.