Counted by hand. Once a week. Already wrong.
Every bar runs on a number it cannot trust. Inventory is counted by hand, once a week, hours of labor producing a snapshot already days stale by the time anyone reads it. In the gap between counts, margin leaks, and the operator living inside it has had no way to see the loss while it is happening. The loss is not theft alone. It is structural, and it is invisible until it is gone.
A single venue recovers ~$33,000 a year
The honest way to size this problem is from the bottom up, at the venue, not from a macro headline. On a venue doing $200K in annual liquor revenue, the recoverable loss breaks into three measurable buckets. This is the figure that actually sells the system, because an operator can check it against their own books.
From -35% shrinkage, Gaming first-year actual.
From -55% backstock, capital freed from shelf.
From 4 to 6 hours per week of manual counts.
The shrinkage and backstock percentages are Gaming first-year operating actuals, single venue type, scoped as such. The $7,000 labor figure and the $200K reference venue are illustrative targets, not measured across a venue population [pending source verification].
The gap between when loss happens and when anyone sees it
The loss compounds in the gap between counts. An extra half-ounce per pour does not look like theft and never shows up in a weekly count, but across a busy bar it is thousands a month in margin that simply evaporates. By the time a manager reads the number, the damage is days old and unrecoverable.
An over-pour, a missed transfer, a spill. No one sees it.
It repeats every shift, invisibly, against no live signal.
Hours of labor produce a stale snapshot of last week.
The number names a loss already gone. Nothing can be done.
That gap is the entire reason Klynkz exists. A real-time read closes it: loss is detected as it happens, while a manager can still act on it, rather than reconciled days later when it is already a write-off.
The real competitor is the clipboard
The honest competitive frame is not a rival app. It is the do-nothing manual count, the clipboard or spreadsheet a venue already runs for free. That is what Klynkz actually displaces, and it is why the per-venue ROI matters more than any feature war.
| Property of the manual count | What it costs the operator |
|---|---|
| Slow | 4 to 6 hours of skilled labor every week, repeated indefinitely. |
| Subjective | Eyeballing a bottle's fill level is a judgment call, not a measurement. |
| Batch | A weekly snapshot, blind to everything that happened between counts. |
| Labor-heavy | The count competes with running the venue, so it is rushed or skipped. |
The category frame that survives diligence: operators commonly lose on the order of 20 percent of profit to shrinkage, and an acceptable shrinkage rate is generally cited as under 2 percent of inventory (trade and industry sources, e.g. Sculpture Hospitality, WISK). That is the citable anchor. The answer to "we already do it ourselves for free" is the $33,000-a-year recovery, not a feature.
The macro figure, scoped honestly
The category is large, and the market story should survive the questions a sophisticated investor asks rather than invite them. The defensible anchors are sourced; the often-quoted $55B figure is presented as context pending a named source, never asserted as fact.
| Figure | What it is | Status |
|---|---|---|
| ~696,000 establishments | US food-and-drink establishments (BLS NAICS 722) | Government data. Category ceiling, not addressable base. |
| ~20% of profit | Typical share lost to shrinkage; under 2% acceptable | Trade and industry sources. Citable. |
| ~$162B | US restaurant food-waste cost, USDA-derived | Secondary, USDA basis. Broader than shrinkage. |
| $55B / 22% rate | Industry shrinkage figure quoted in beverage material | [pending source verification]. Context, not asserted. |
The ROI case does not depend on the $55B number at all. It depends on the per-venue $33,000 recovery, which is the figure that actually sells. The macro figure is presented only as context, and only with its source status shown.
Why now
Margins in hospitality are under sustained pressure, and the labor that ran the weekly count is scarcer and more expensive than it has ever been. A tool that pays for itself many times over while removing hours of skilled labor is a clearer purchase in 2026 than it would have been a decade ago.
The enabling technology also only just became viable. Reading RFID reliably near liquids required the concentrating-layer physics that Klynkz patented in 2023. The full mat-to-cloud system was issued in 2022. The reason no incumbent built this is not that no one wanted it. It is that the physics and the IP did not exist until Klynkz created them.