Short answer: On-premise and off-premise are not variations of the same game — they are structurally different channels that require different data, different success metrics, and different investment logic. Brands that apply a single analytics framework to both consistently misread performance and misallocate sales resources.
Why the Same Metrics Don’t Work for Both Channels
Off-premise (retail) favors volume, velocity, and shelf share. The question is how fast product moves through a SKU in a defined set of stores. Data is relatively available — retail POS, scanner data services, and store-level shipments provide a reasonably clear picture.
On-premise (bars, restaurants, event venues) operates on different economics entirely. Volume per account is lower, but brand-building value is higher. A listing at the right high-visibility restaurant can influence consumer trial and subsequently retail pull. The data environment is also structurally murkier — there is no consistent on-premise POS aggregator for the drinks trade at the regional brewery scale.
Non-alcoholic beer adds further complexity. NA beer is disproportionately indexed toward on-premise in its early distribution phases, because restaurants and bars have a specific need for credible alcohol-free options on their menus. The brand-building logic of on-premise placement is particularly strong for NA — consumers discover NA options in social settings where they can try them without the commitment of a retail purchase.
The Off-Premise Intelligence Framework
Off-premise analytics for beverages centers on three questions:
1. Velocity per point of distribution (POD): Total depletions divided by active retail accounts in the period. This normalizes volume for distribution breadth — a brand in 50 stores selling 200 cases is outperforming one in 200 stores selling 300 cases, and velocity per POD makes that clear.
2. Shelf share vs. competitive set: What percentage of linear shelf space or SKU facings does your brand hold relative to the category? In chain retail, shelf resets create structural opportunities and risks that velocity data alone doesn’t capture.
3. Promotional lift and efficiency: When price promotions or display programs run, how much incremental volume do they generate? Brands that can quantify promotional ROI by account tier gain significant negotiating leverage in retailer planning conversations.
The On-Premise Intelligence Framework
On-premise analytics requires different proxies because direct POS data is rarely available:
1. Menu penetration rate: What percentage of target on-premise accounts carry your brand on their menu or tap list? Track this through rep visit logs and periodic audits. For NA beer, track placement in the non-alcoholic section specifically — presence there signals deliberate category development by the operator.
2. Draft handle placement and tenure: For draft products, track handle count (how many accounts have a dedicated tap) and tenure (how long each handle has been active). A handle that stays active for multiple quarters is a meaningful signal of genuine pull. Short-tenure handles that turn over rapidly suggest a trial-and-reject pattern worth investigating.
3. Staff advocacy index: In higher-investment on-premise accounts, track whether floor staff can accurately describe and recommend the product. This is a qualitative metric gathered through rep observation, but it is a leading indicator of sustainable velocity in experience-driven venues.
Building One Integrated View
The channels are analytically distinct but commercially interdependent. A practical integration point: use on-premise depletion velocity and menu penetration in high-visibility accounts as a leading indicator of off-premise demand in the same geography. Brands often see off-premise velocity uptick in ZIP codes where on-premise presence has been built deliberately — and can use that signal to time off-premise expansion.
See also: Account Scoring: Finding Your Next 100 Outlets for how channel type feeds into account-level prioritization.
Where Channel Intelligence Breaks Down
- On-premise data is genuinely hard to get right. Distributor depletion data at the account level is the best available proxy, but distributor reporting quality for on-premise accounts is often less reliable than for retail. Treat on-premise depletion data as directional, not precise.
- Channel attribution is messy. Proving that on-premise investment drove off-premise lift requires controlled market analysis that most regional breweries cannot practically execute. Treat the relationship as a working hypothesis informed by market observation.
- NA beer on-premise benchmarks are still being written. What a “good” NA beer menu placement looks like in terms of velocity, tenure, and staff adoption is an open question. Document your own experiences systematically — your data will compound in value as the category matures.
Part of the Sales Intelligence track — browse all.
Frequently asked questions
What is the key difference between on-premise and off-premise data? Off-premise data (retail) is more structured and available — scanner data, shelf share reports, and retailer POS are relatively accessible. On-premise data (bars, restaurants) is fragmented and largely self-reported, making it harder to aggregate but often more valuable for brand-building signals.
Which channel matters more for launching a new beer or NA beer brand? On-premise is generally the stronger brand-building channel for new launches — accounts provide sampling opportunities, staff advocacy, and visible brand presence that drives off-premise trial. However, off-premise volume is where brands sustain revenue once trial is established.
How do you measure on-premise performance without reliable POS data? Proxy metrics are the practical alternative: depletion velocity at the account level (from distributor data), menu placement (tracked through periodic audits or rep visits), and pour handle count for draft products. None are perfect, but together they approximate the performance signal that POS data would provide directly.