If you sell beverage alcohol through distributors, you don’t see your own sales. You see depletions. That distinction trips up nearly every analyst who comes into the industry from regular CPG, and the public internet is almost no help. The freshest substantive explainer we could find when we went looking was a 2021 Overproof post on depletion reports, and a 2016 press release. So here’s the 2026 version, written from the data-engineering side.
We’re gmware, a software and data engineering firm in Austin, TX, with delivery centers in Bangalore and Mohali, India. We also run Shield Suite, our own retail-intelligence platform for beverage-alcohol brands across 60,000+ storefronts. That means we spend our days inside exactly the systems this post is about (VIP extracts, iDIG logins, distributor portals that all name the same product three different ways) and we’ve reconciled enough of them to know where the bodies are buried.
This is a primer for the brand analyst, the new distributor BI hire, or the founder of an emerging label who just got told “we’ll send you the depletion report” and isn’t sure what that report can and can’t tell them.
| Term | What it is in one line |
|---|---|
| Depletion | A case shipped from distributor warehouse to a retail or on-premise account |
| VIP / iDIG | The systems distributors run on (capture) and brands log into to read it (surface) |
| SipSource | The WSWA’s national aggregation of distributor depletion data |
| Scan data | Retail register sales from syndicated panels (NIQ, Circana) |
| Store-level POS | Raw transactions from an individual store’s register |
What depletion data actually is
A depletion is a shipment out of a distributor’s warehouse to a downstream account. That’s the whole definition, and the whole problem. Under the U.S. three-tier system, most suppliers legally can’t sell straight to a store. They sell to a distributor, who sells to retail. So the supplier never touches the point of sale. The deepest demand signal they get is the moment a case leaves the distributor’s dock.
That makes depletions a leading indicator with a built-in lie. When a chain buys three months of holiday inventory in October, depletions spike in October even though nothing sold to a single shopper yet. Depletions tell you what moved through the middle tier. They do not tell you what a human being bought. Hold onto that gap, because every downstream confusion in beverage-alcohol analytics traces back to it.
How VIP, iDIG, and SipSource fit together
They’re three layers of the same pipe, not competitors. The cleanest way to keep them straight is capture, surface, aggregate.
VIP builds the operational software a large share of U.S. distributors actually run their warehouses and sales teams on, so it sits at the point where depletion data is captured. iDIG is the analytics front end, the place brand and supplier analysts log in to surface and slice that captured data against their own portfolio. (VIP and iDIG now sit under the same Karma Notes / iDIG umbrella after years of industry consolidation; if you’re integrating, treat them as one ecosystem.) SipSource, run by the Wine & Spirits Wholesalers of America, sits on top and aggregates distributor depletions into national and regional trend reports, the numbers you see quoted in trade headlines about category growth.
So when someone says “pull the depletion data,” ask which layer. The distributor’s raw extract, the iDIG view, and the SipSource trend report are three different grains of the same underlying flow, and they will not tie out to each other on the first try.
Capture, surface, aggregate
Depletions vs scan vs store-level POS
Each data source answers a different question, and the expensive mistakes happen when a brand asks one of them a question it structurally can’t answer. Here’s the honest version.
| Source | Measures | Strength | Blind spot |
|---|---|---|---|
| Depletions (VIP/iDIG/SipSource) | Cases shipped distributor to account | Earliest demand signal; full distributed footprint | Not retail sales; no shelf, price, or sell-through |
| Syndicated scan (NIQ, Circana) | Register sales in a measured panel | True consumer takeaway; category context | Thin independent-store coverage; subscription cost |
| Store-level POS | Raw transactions at one store | Exact price, basket, sell-through | Only the stores you have a feed from |
Which source answers what
Syndicated panels from NIQ and Circana measure what actually rings up at the register: real consumer takeaway, which depletions can’t give you. The catch in alcohol specifically is independent-store coverage. A large share of off-premise beverage-alcohol volume runs through independent liquor stores that syndicated panels measure thinly or not at all, so a brand that over-indexes on independents can look flat in scan while depleting fine. Store-level POS fills that hole, but only for stores you’ve struck a data relationship with. Almost every serious brand ends up blending all three.
Why depletions and retail sales never tie out
Because they’re measuring two different events separated by time and a shelf. Depletions record the case leaving the distributor; retail sales record the bottle leaving the store. Between those two moments sits inventory, and inventory absorbs the difference.
The classic reconciliation failure is the buy-ahead. An account loads up before a holiday or a price increase, depletions jump, and the brand manager celebrates a demand surge that’s really just a warehouse moving its restocking forward a few weeks. Run sell-through analytics on that period and you’ll see flat consumer takeaway under a depletion spike. The reverse happens too: a hot product sells through fast, the account is out of stock, and depletions go quiet not because demand died but because there’s nothing left to ship. If you treat depletions as sales, you’ll chase both ghosts. This is the same shelf-level reality our liquor store analytics guide covers from the retailer’s side of the counter.
Integrating depletion feeds into your own warehouse
The pipe is the easy part. Pull the iDIG export or distributor extract on a schedule, land it raw and immutable, then transform. We can wire that flow up in days. The reason these projects take months anyway is the master data underneath.
Across a national footprint, the same SKU carries different distributor item codes in different markets. The same chain shows up under five spellings of its name. Reporting calendars don’t agree: some distributors close their month on a Saturday, some on the last day, some on a 4-4-5 retail calendar. None of that is exotic; all of it has to be resolved into a shared product master and account master before a single dashboard number is trustworthy. The first time we scoped one of these, we under-budgeted the mapping line and learned the lesson the hard way. Plan for the reconciliation, not the connector.
- Land raw first. Keep an immutable copy of every extract before you transform, so you can always re-derive when a mapping changes.
- Build the product and account masters before the dashboard, not after. Retrofitting a master onto live reports is the painful path.
- Reconcile depletions against scan or POS on a fixed cadence so buy-ahead distortions surface instead of hiding.
- Version your distributor calendars so a month-close difference doesn’t read as a demand swing.
How to wire depletion feeds in
This is squarely the kind of work our data analytics and business intelligence and big data consulting practices do for beverage-alcohol clients, and it’s the foundation every forecasting model sits on top of.
When an emerging brand should bother with depletion data
When you have enough distribution that you can’t call your accounts to ask how things are moving. Below that line, depletion infrastructure is overkill. A sub-$5M label with a handful of distributor relationships will get further pulling the free iDIG views their distributors already provide and tracking them in a spreadsheet than standing up a warehouse.
When to invest, three converging signals
The trigger to invest is when three things converge: distribution spreads across enough markets that manual tracking breaks, the distributor product and account naming starts diverging, and you begin caring about sell-through rather than just shipment volume. That’s usually also the moment you want to start blending depletions with whatever store-level POS or syndicated data you can get, because shipment data alone, as we keep saying, leads the truth and sometimes lies about it. Spend the money when you’ve outgrown the phone call, not before.
What we’d recommend
Don’t buy a depletion-analytics platform before you’ve decided what question you’re trying to answer, because the answer changes which data you actually need. If you want the earliest read on demand across your whole footprint, depletions win and iDIG is where you’ll live. If you want to know whether product is selling through versus sitting, you need scan or POS layered on top, because depletions can’t see the shelf. Most brands we work with need both, and the real engineering project is the reconciliation layer that lets them trust the blend.
When we build these for clients, the same dual-shore model behind Shield Suite applies: onshore architecture and account oversight out of Austin, with the heavier data-engineering build run by our India delivery teams during overlapping hours. It keeps the reconciliation work, which is most of the work, at sane economics. If you’re moving from a single-distributor spreadsheet toward a real warehouse, our AI demand forecasting guide for distributors is the logical next read, since clean depletion data is what makes forecasting possible at all.
Tell us what your data flow looks like today (which distributors, which feeds, how messy) and we’ll give you a straight answer on scope, cost, and timeline within 48 hours. We work with beverage-alcohol brands and distributors on exactly this problem.