Understanding key differences between POS and panel data

By Chelsea Stone

8 min read

January 5, 2021

The number one challenge we hear from small businesses is not that the CPG industry lacks data – it’s that there is too much data available. If you’re a small CPG manufacturer trying to tackle the foundations of retail data, it’s hard to know where to begin. Learning the CPG retail data landscape can feel like studying for a doctoral degree, but it doesn’t have to be so complicated. The best place to start is understanding the foundations of your data – what can POS (point-of-sale) data tell you? And what can panel (household) data tell you?

What kind of questions can POS vs panel data answer?
POS (retail measurement) data
Panel (household, or shopper) data


POS vs panel data collection

As CPG industry professionals, we know that we need both POS (point-of-sale) data as well as panel (household, or consumer-level) data in order to make the best decisions for our business.

POS data is what’s collected at the time of sale. You may recognize the word “POS” as the electronic system we often call a register. Every bar code the cashier scans can tell us something – the price, quantity sold, time of sale, whether a product was purchased on promotion, what retailer the product was purchased at, etc. Panel data, on the other hand, comes from Nielsen’s panel of households who provide their purchasing data by scanning the barcodes on products they purchase and indicate where they bought it. This allows Nielsen to track consumer behavior for more than 250,000 households in 25 countries, and grants CPG data professionals insights like customer loyalty, consumer demographics, and the purchase cycle of a given product.

With so many metrics available to us from looking at POS and panel data, how do we know what to focus on to derive the most value in an industry in which we need quick, simple answers now?

A streamlined look at the value and purpose of POS and panel data can point us in the right direction.


What questions can POS vs panel data answer?

Sales can be “decomposed” based on POS or panel metrics. While a POS decomposition will focus on the tactical levers of “The 4 P’s” (Price, Promotion, Product, and Place), a panel decomposition will focus on who buys your product and how they shop.

POS data

POS data, also referred to as retail measurement data, can provide a wealth of insights. The easiest way to begin is by having a look at a Byzzer report called Promo versus Non-Promo Decomposition Tree:

Whether your sales are up or down, looking at the “decomposition” of your POS data (i.e., those factors that could drive changes in your total retail sales), will help identify possible sources of a change in sales.

Non-promo sales versus promo sales
Non-promo sales refer to product sold at your everyday price, while promotion sales refers to merchandising, and includes product sold that was on a display, in a feature, or offered at a discount.

Non-promo sales
One helpful way to examine the sales of your product at everyday prices is to take a closer look at distribution and velocity of sales. These two are directly related to each other – for folks interested in the math, your sales equal your distribution multiplied by your velocity. They work hand-in-hand because distribution is how widely available your product is, versus velocity, which is how quickly your product sells where it’s available. Using this visual, you can see if your product has lost or gained shelf space, or if sales are slowing or gaining steam.

On their own, these measures can give you a topline overview of how your product is selling. To dig deeper for a diagnosis, you’ll want to look underneath the hood of your velocity and distribution. Remember, velocity equals sales divided by distribution, so if your velocity has dropped, could it be a problem of distribution? If so, distribution where? In a single city or state? In a region? In a retailer? You’ll want to take a closer look using a Byzzer report like Market Opportunity.

Promo sales by type
Display: Ever been enticed by a product because it was displayed at the end of an aisle or in its own special little section at the grocery store? These kinds of secondary displays are one way to drive promo sales.
Feature: A feature could comprise your product being included in the retailer’s weekly flyer or circular, often featured in a retailer’s email campaigns.
Discount: A temporary price reduction (often referred to as TPR, as in the chart above) can help drive promo sales, too.

Note that any one or any combination of these three efforts constitute a promotion, so a product could be on its same shelf as usual on a discount, or it could be on a display – and priced the same, on discount, or even higher than usual.

Incremental versus non-incremental sales
Congratulations – you implemented a promotion and your sales went up! It’s because you did such a great job on your promotional activity, right? Not necessarily. Incremental versus non-incremental sales will tell you how much product you sold due to the promotion itself (incremental) versus how much product would have sold anyway had you done no promotion at all (non-incremental).

To dissect your promotional performance, trend your promotional performance with the Byzzer report Promotional Incremental Sales Trend or check out your subsidized volume by market with the Byzzer report Promotion Efficiency Comparison.


What about panel (household) data?

Comparing penetration to spend per household is a little bit like comparing distribution to velocity in your POS data. Sales trends can be due to either a change in how many households are buying the product (penetration) or current buyers spending less or more ($ per household). Declining penetration indicates that you are losing share of households. Dollar per household changes could indicate a change in category usage, and may be driven by shoppers buying the product on fewer trips (frequency) or spending less per trip ($ per trip). That means, for example, that panel data can give us insights like: “The average household spent $100 on dog treats in 2020 across 5 trips with an average spend of $20 per trip.”

Using the decomposition above, we could identify problems and formulate potential solutions with the right marketing strategy. For example:

If penetration has decreased, a CPG manufacturer could try new customer acquisition tactics, such as influencer marketing or programmatic ads.

If frequency of trips has dropped, a manufacturer could deploy discounts instead of Buy-1-Get-1 deals.

If dollars spent per trip have dropped, a manufacturer could implement “must-buy” deals such as 2 for $4.

A closer look at Byzzer’s reports can inspire which potential action to take – and you’ll be able to justify any new marketing initiatives by showing the size of an opportunity. Start with Shopper Sales Decomposition Tree to see what you can glean from topline panel data intel, and move on to a report such as Shopper New, Lost and Retained or the Leakage Tree to pinpoint where you should take action.


Start with a free alert and report today.

See your products in more carts.

Get free access

Byzzer Content Hub

The pet food trends and attributes that drove in-store sales and online searches in 2020-2021

5 minute read

According to the 2021-2022 APPA National Pet Owners Survey, 70% of US households own a pet. But for an increasing...

What CPG brands can do to capture growth from strong trends in plant-based protein?

5 minute read

Go into any average grocery store today, and you’ll likely see Beyond Burgers alongside traditional beef patties and almond milk...

How well does your brand leverage price?

5 minute read

Setting product prices is a conundrum that affects all manufacturers. You think you know how to weigh your production costs...

Stay in the loop with industry insights.

We’ll send you the latest and greatest on Byzzer, CPG, retail, manufacturing, and more.