How to Decompose CPG Sales Using Syndicated Data

When a CPG brand has a strong quarter, everyone is pleased. When a brand has a bad quarter, everyone wants to know why. In both cases, the answer is the same: you need to decompose the sales.

Sales decomposition means breaking total volume change into its component parts. Which portion of the growth came from distribution gains? Which came from velocity improvement? Which came from pricing? Which came from promotional activity?

Without decomposition, CPG teams are guessing. They see a number go up or down, and they build a narrative around it. That narrative may be directionally right. But it is not the same as knowing. And in a buyer meeting, the difference between knowing and guessing is visible.

This post walks through how to decompose CPG sales using syndicated data, where the standard approaches break down, and what gets lost when teams skip this step.

The Core Components of CPG Sales Decomposition

At the top level, CPG sales can be broken into four drivers. Each one can be isolated using syndicated data from SPINS, NielsenIQ, or Circana.

Distribution Contribution: How much of the sales change came from gaining or losing points of distribution? A brand that grew total sales 15 percent but also added 20 percent more shelf placements is not growing on a per-store basis. Distribution contribution isolates the volume driven purely by more (or fewer) locations.

Velocity Contribution: How much of the change came from selling more per store that already carries the product? This is the metric that tells you whether your brand is actually winning on shelf — not just showing up. Velocity is measured as dollars or units per point of distribution, per week.

Pricing and Mix Contribution: How much of the dollar change was driven by price per unit, rather than unit volume? A brand that raised prices and held units roughly flat has a very different story than a brand that cut prices and grew volume. Without separating price from volume, dollar sales growth can look the same across two brands with completely opposite trajectories.

Promotional Contribution: How much of the volume spike during the period came from promotional events? Promo lift needs to be isolated from base velocity to understand whether underlying demand is growing or whether the brand is becoming dependent on trade spend to hit its numbers.

How to Run a Sales Decomposition in Syndicated Data

The mechanics start with a period-over-period comparison — typically four weeks, 13 weeks, or 52 weeks. For each period, you need:

• Total dollar and unit sales by retailer and by item
• ACV-weighted distribution by item
• Velocity (dollars or units per point of distribution per week)
• Average selling price and promotional price depth
• Promoted versus non-promoted weeks

With those inputs, you can calculate the contribution of each driver to the total change:

• Distribution effect = Change in ACV distribution x Prior-period velocity
• Velocity effect = Change in velocity x Current-period distribution
• Price effect = Change in average selling price x Current-period unit volume
• Promo effect = Promoted week volume minus baseline volume, summed across the period

The four effects should sum to approximately the total sales change. Differences between the sum and total are typically explained by interaction effects — the compounding between distribution and velocity changes in the same period.

Where Standard Syndicated Outputs Fall Short

Most syndicated data platforms give you total sales and distribution metrics. They do not pre-build decomposition. That means the analyst has to construct it manually from exports — pulling the right hierarchy levels, aligning time periods, and calculating the driver contributions in a separate model.

Three gaps show up consistently in practice:

First, retailers are not separated. A total US or xAOC decomposition tells you the national story but obscures what is actually happening by account. A brand that is growing distribution at one major retailer but losing velocity at another needs two decompositions, not one blended view. Account-level analysis requires building it separately for each retailer.

Second, promotional weeks distort the decomposition unless they are explicitly modeled. If the measurement period contains a major promotional event, the velocity contribution will look inflated relative to a non-promoted period. Without stripping out the promo effect first, the decomposition overstates organic velocity improvement.

Third, SKU-level decomposition gets lost in aggregation. A brand-level decomposition might show stable overall velocity while one high-velocity SKU is declining and another is growing to compensate. The portfolio-level number masks the rotation happening underneath.

What Good Decomposition Looks Like in a Buyer Meeting

The goal of sales decomposition is not the analysis itself. It is the conversation it enables.

When you can walk into a category review and show a buyer that 60 percent of your growth was driven by velocity improvement (not just distribution gains), you are making a different argument than a brand showing raw dollar growth. Velocity improvement is a signal that the product is resonating with the buyers’ shoppers. Distribution gains can be temporary. Velocity trends are what buyers watch to decide whether a brand deserves more shelf space.

The brands that earn more shelf space are not always the ones with the highest dollar sales. They are the ones who can explain exactly why their sales moved the way they did.

How Bedrock Makes Decomposition Faster

Bedrock’s analytics platform runs harmonized data across SPINS, NielsenIQ, Circana, and retailer portals in one place. Because distribution, velocity, pricing, and promotional data all live in the same data layer, decomposition runs at the account and item level without manual export and reconciliation.

The insight — which driver is contributing most to the change — surfaces automatically alongside the numbers. Teams spend time acting on the decomposition instead of building it.

See how Bedrock surfaces sales decomposition across your retailer portfolio. Book a demo or visit bedrockanalytics.com.