A busy hotel bar can generate a full week of reports without answering the questions that drive margin: what to order, what to batch, where variance leaks, which outlet needs training, and which menu items deserve shelf space. That gap, according to beverage manager and consultant Riccardo Grechi, is not a technology shortage. Hotels already capture sales in the point-of-sale system, purchases in inventory tools, theoretical consumption in recipe files, labor in scheduling platforms, and waste and transfers somewhere else entirely. The problem is that these records describe different versions of the same service.

In an analysis published by Hotel Technology News, Grechi — founder of The Double Strainer, a consultancy focused on menu engineering, costing and operational execution, with more than 10 years across luxury hotels and independent venues in Europe and Asia — argues the operational opportunity is not another dashboard but a decision system that connects sales, stock movement, recipes, exceptions and service conditions closely enough for managers to act while the information still matters.

Start With the Decision, Not the Data Fields

Most reporting projects begin with a list of available data fields. Grechi flips that: start with the recurring decision an operator must make, then define the action, the owner and the frequency. He identifies five decision categories where beverage data should earn its keep:

• Purchasing: how much stock to order for each outlet, event and demand period.
• Production: which batches, garnishes and syrups to prepare, in what quantity and when.
• Menu: which drinks combine strong demand, contribution margin and operational feasibility.
• Staffing and training: where slow execution, inconsistent portions or repeated errors point to a skills gap.
• Control: which variances warrant investigation and which are explained by approved transfers, waste or comps.

Once the decision is fixed, the reporting requirement shrinks. A daily production plan needs recent sales patterns, reservations, events and current stock. A monthly menu review needs sales mix, contribution margin, prep burden and recurring variance. Merging every metric onto one screen, Grechi writes, creates visibility without direction.

One Version of Every Drink

Connected reporting depends on a shared unit of truth: the active recipe. Each sellable item should carry one current recipe, one defined yield, one POS identifier and a clear link to inventory units. That sounds basic, but hotel beverage programs change constantly. A spirit gets substituted while purchasing waits on a delivery. A garnish disappears during a busy weekend. A batch recipe is adjusted while the inventory system still runs the old yield. Banqueting pours a different portion than the lobby bar. Each change widens the gap between theoretical and actual consumption.

Version control, Grechi argues, matters more than sophisticated analytics when the underlying recipe map is unreliable. Without a record of when a recipe changed, who approved it, which outlets use it and how revised quantities affect theoretical cost, automated variance reports simply calculate discrepancies faster.

Five Signals, Not One Metric

No single number explains beverage performance. Operators need five connected signals, and the value comes from comparison. If actual depletion exceeds theoretical usage and exceptions go unrecorded, the hotel has a control or data-capture problem. If usage matches but contribution margin is weak, the issue is price, recipe cost or sales mix. If a profitable item is frequently unavailable, production planning — not demand — is capping revenue.

Pour cost in particular falls short as a decision rule. A low-cost cocktail that sells rarely, takes minutes to build and generates waste may contribute less than a higher-cost drink with strong demand, fast execution and reliable availability. Grechi recommends reviewing menus across at least five dimensions: contribution margin per sale, sales velocity, preparation and service burden, ingredient variance, and availability during the hours guests actually want the item.

Outlet-level segmentation adds another layer. The same beverage performs differently in the lobby bar, pool bar, club lounge, room service and banqueting because demand patterns, staffing and service expectations differ. A high-volume conference week should not set the purchasing pattern for an intimate cocktail bar; a pool bar's weather sensitivity should not distort a lobby lounge forecast. Property-wide totals can hide that operational story.

Capture Exceptions During Service

Variance is often read as evidence of poor control, but unexplained variance and operational variance are not the same. Tastings, guest recovery, spills, breakage, recipe trials, inter-outlet transfers and substitutions can all be legitimate. Recorded late, inconsistently or on paper, the system loses the reason behind the movement. The best exception process is short enough to use mid-service — item, quantity, reason, outlet, time and author, logged without leaving the workflow. Ten separate "minor" waste entries tied to the same garnish may reveal a production problem a monthly stock count cannot explain.

An Operating Rhythm

Data becomes operational when it has a cadence. Before service: review reservations, occupancy, group business and stock constraints, and adjust prep and transfers before teams get busy. After service: capture shortages, waste and substitutions while the shift is fresh. Weekly: compare sales mix, theoretical usage, actual depletion and high-value exceptions, and assign corrective actions to a named owner. Monthly: review menu performance, supplier changes, recipe versions and the effect of promotions. A report without an owner, Grechi writes, is an archive; each review should end with a decision, a responsible person and a date to check the result.

What to Demand From Technology

Hotels do not need one system doing everything; they need reliable information moving between the systems they already run. Grechi's checklist: consistent identifiers and recipe mapping across POS, inventory and costing; outlet and occasion segmentation without losing the property view; time-stamped version control on recipe and price changes; accessible exception capture at the point of activity; usable exports and integrations; and role-based visibility so executives, finance and outlet managers work from the same underlying records.

A Thirty-Day Pilot

A hotel can improve beverage decisions without replacing its tech stack. In week one, choose one decision to improve in one outlet — ordering, production or menu review. In week two, verify active recipes, yields, POS identifiers and inventory units for the outlet's highest-volume items. In week three, connect sales, theoretical consumption, actual depletion and exceptions for those items. In week four, run one review meeting, assign actions and document what information was missing. The pilot surfaces the real integration problem before the hotel commits to a larger reporting project — and produces a repeatable workflow rather than a dashboard that looks complete but changes nothing.

The goal, Grechi concludes, is a credible signal early enough to adjust an order, modify a batch, retrain a team member or reposition a menu item before the next service. When technology, recipes and operating routines describe the same reality, beverage data stops being a reporting exercise and becomes part of service management.