CPG Analytics
CPG analytics
CPG analytics is the work of turning a brand’s sell-through, distribution, and trade data into decisions: which SKU to push, which promotion paid back, where distribution is leaking. This guide covers the data it runs on, the core metrics, what marketing analytics adds, and how to choose the tools.
What CPG analytics is
Every consumer-packaged-goods brand already sits on more data than it uses. The problem is rarely access. It is that the data disagrees with itself: retailer portals, syndicated panels, and the trade-spend spreadsheet each describe the same business in a different item hierarchy, a different week, and a different unit, so the week before any real analysis is spent making them line up.
CPG analytics is the discipline of doing that reconciliation once and then asking useful questions of the result: is this velocity drop demand or distribution, did this promotion clear its cost, which retailer is the growth coming from. The mechanics are specific to the channel, which is why a general analytics tool struggles here. It does not know what an ACV point or a syndicated week means until someone teaches it.
The data CPG analytics runs on
Four data types, four different shapes. The hard part of CPG data is never one source. It is making all four agree on what a unit and a week are.
Sell-through (POS) data
Units and dollars scanned at the register, by store and week. The ground truth of demand: what a shopper actually bought, not what you shipped. It arrives from retailer portals like Walmart Retail Link and from syndicated panels, on different hierarchies that have to be reconciled before they agree.
Shipment and distributor data
What left your warehouse and what the distributor pulled. Useful for supply and trade reconciliation, but a lagging, lumpy proxy for demand: a big shipment week is often a buy-in ahead of a promotion, not a demand spike.
Syndicated market data
SPINS, NielsenIQ, Circana. Your share and your category, the only view that includes competitors. Priced per category and delivered on its own schedule, which is why it usually lives in a separate deck from your own numbers until something forces them together.
Trade and promotion data
What promotions cost (off-invoice, billbacks, manufacturer chargebacks) and what they returned in incremental lift. The most expensive data a CPG brand owns, and the one most often stranded in a finance spreadsheet no analyst ever joins to sell-through.
The core metrics
Most CPG decisions come back to four numbers. Computed correctly they are decisive; computed loosely they mislead in confident ways.
Velocity (units per store per week)
The single most-used CPG metric: how fast a SKU sells where it is carried. It normalizes for distribution, so you can compare a product in 200 stores against one in 2,000 without the store count drowning the signal.
ACV-weighted distribution and TDP
How much of the market can actually buy your product, weighted by store size rather than store count. A SKU in 100 large-format stores can outreach one in 400 small ones. Total Distribution Points (TDP) folds depth into the same number.
Baseline vs. incremental lift
Splitting what would have sold anyway from what the promotion drove. Without this split, every promo looks successful because total units went up. The question is whether the units above baseline covered the discount that bought them.
Distribution gains and losses
Points of distribution won or lost, by retailer and division. A velocity drop that is really a distribution loss is a different problem with a different fix, and conflating the two is the most common misread in a category review.
CPG marketing analytics
CPG marketing analytics is the same toolkit pointed at spend instead of shelf. The questions shift to attribution: did the shopper-marketing program at a banner lift velocity above baseline, did the price-pack change hold margin without bleeding units, did a competitor’s promotion steal the week.
The limit is that retail data is observational, not an experiment. You rarely get a clean control store, so marketing analytics in CPG leans on baseline-versus-incremental decomposition and like-for-like comparison rather than true causal lift. Treat a tool that promises clean ROI on every tactic with suspicion. The data underneath usually cannot support the claim.
Spreadsheets, BI tools, and CPG data tools
Spreadsheets carry a brand further than vendors admit. With one or two retailers and a disciplined analyst, a workbook is the right tool, and buying software early just adds a subscription to a problem that was not yet expensive.
General BI tools like Tableau, Power BI, and Looker chart whatever you load but carry no retail knowledge. The analyst still harmonizes the feeds and defines every metric by hand, so the tool speeds up the last step and leaves the slow one in place.
Purpose-built CPG data tools earn their place when the retailer count or the trade-spend complexity crosses the line where the reconciliation itself is the bottleneck. The test is the same one every time: if the team spends more of the week preparing CPG data than interpreting it, the specialized tool pays for itself in recovered analyst hours before it does anything clever.
Where Scout fits
Scout is a CPG analytics platform built for brand teams. It ingests the data above (portal exports, POS files, SPINS and other syndicated panels, and the trade-spend stream), harmonizes the item codes and week definitions, and computes velocity, distribution, and lift in one place. The AI layer lets an analyst or a category manager ask a question in plain language and get a chart back instead of filing a request.
The honest boundary: Scout is a demand-side analytics layer. It is strongest on sell-through, distribution, and trade-promotion analysis, not on the supply planning, ERP, or claims-recovery work that other systems own. It reads the data those systems produce; it does not replace them.
Related: Retail intelligence platform · CPG software · Syndicated data · Sales cannibalization
Tell us what you’re working on
A 30-minute conversation to scope fit. Pick a time that works for you.