ACNIELSEN INSIGHT

Using Panels to Understand the Consumer
by Ken Greenberg, Vice President, Marketing, ACNielsen Homescan, US.

DOWNLOAD THIS ARTICLE
CLICK HERE

Email this Article to a Friend

The Consumer Is King

Understanding consumers and what drives their purchasing behaviour is one of the most basic building blocks for developing a successful marketing strategy.  The diagnostic tool that provides this powerful information is household panel data.  Panel information, unlike scanning data, is used to explain the reasons behind period-to-period volume changes. 

This overview of how consumer panel information can be used in a marketing plan is intended to help you use panel data to the fullest potential.  This article will address: the definition of household panel data; the differences between scanning data and panel data; the key components of panel data that allow you to analyse and understand volumetric data; and the basic behavioural measures that consumer panel data delivers. 

What Is A Household Panel?

Household panel data is collected after each panelist shopping trip.  Members of the panel record their purchases, capturing not only what is purchased, but also where the purchase was made (store or channel), and whether the purchases were the results of a promotional deal.  This purchase information is then tied back to the general demographics of the household. 

Household panel data allows marketers to measure the ongoing purchase habits and practices of household and demographic groups.  Tracking and analysing this information over time can reveal the dynamics of consumer purchasing – such as who is buying your product, how often they buy, the competitive products your buyers are also purchasing, and whether buyers are deal sensitive.  Panel data quantifies the composition of category or brand volume, which can then be used to identify the appropriate marketing strategies to drive growth or defend against competitive actions.  Key issues such as consumer purchase behaviour, shopping habits, demographics, attitudes and opinions are critical elements the panel can deliver. 

Panel data quantifies the composition of category or brand volume. 

The ACNielsen Homescan consumer panel of 61,500 US households collects consumer shopping and purchase data from all outlet channels, including grocery, drug, mass and convenience stores.  The panel is geographically dispersed and is demographically balanced so the sample profile matches the US population as closely as possible.  The panel data is also projected to US-based Census estimates that are updated regularly to reflect population changes.  The Homescan panel is considered by some as the industry gold standard due to its long-standing reputation in the marketplace and its utilisation of revolutionary hand-held technology that changed the face of household panel data collection. 

Each household in the Homescan panel collects purchase information on each shopping trip.  For each shopping trip, the following data is recorded:

·   Date of purchase.

·   Age and sex of primary and secondary shopper.

·   Store name.

·   Usage of frequent shopper cards.

·   Complete item description through UPC dictionary. 

·   For each UPC, the number of units, price paid, and deals used.

·   Dealing – specified by the panellist as manufacturer coupon, store coupon, store sale, or other.

·   Source of the coupon – at home, at the register, elsewhere in the store. 

·   Total shopping trip purchase amount.

·   Method of payment – cash, cheque, credit card, or debit card.

What Is The Difference Between Panel Data And Scan Data?

Scan data is collected at the store point of sale and is the most accurate source of volumetric and or share information.  It addresses the “what happened in the store” questions such as sales volume and share, price, and retail trade support.  Panel data is used to understand the reasons behind volume/share levels and trends.  It answers such key questions as: 

·   Who are your buyers? 

·   How often do they buy? 

·   Where do they shop? 

·   Are they loyal to your brand?

·   How do they respond to your marketing efforts?

Sometimes, marketers question the quality of household panel data when they try to reconcile it with store-scanning data.  There is the perception that the volumetric data from each source should be the same.  However, panel data and store data are not always equal because measurement methodologies differ.  Store-level data records millions of shopping transactions while panel data records a specific group of shoppers.  In addition, panel data only represents household-based purchases, so there are no small businesses or civic organisations included in the panel. 

The bottom line is that both types of information have their uses, and by combining the two, marketers have been able to quantify the composition of volume, understand the reasons behind volume changes, reveal the dynamics of consumer purchasing, and identify appropriate sales and marketing strategies. 

Store-level scanning data can tell you that your sales were down in 2001.  Panel data will provide you with information telling you whether you lost volume due to fewer buyers or if your buyers purchased less.  Panel data also gives you information on which competitors you lost volume to.

Difference Between Panel Data And Scan Data

The Key Measures

There are a number of basic terms that are used in household panel data – penetration, buying rate, purchase frequency, and purchase size.  Brand or category volume is a function of the number of households which make a purchase (penetration) and the quantity that they purchase (buying rate).  Buying rate is a function of purchase frequency (occasions per buying household) and purchase size (quantity per occasion). 

Volume =                 # of Buying Households x Quantity Purchased
                                (% Penetration)                   (Buying Rate)

Buying Rate =          # of Purchase Occasions x Quantity per Occasion
                                (Purchase Frequency)           (Purchase Size)

Let’s take a look at each of these terms individually.  

Penetration is determined by the number of households purchasing the product at least once during the time period.  For example, if there are 10,000 households, and 5,000 households purchase Brand X at least once, then Brand X’s penetration is 50%. 

The buying rate is the average volume purchased by buyers during the given time period.  As an example, if 500 households purchase 3,000 units of Brand X, the buying rate is 6 units. 

Purchase frequency is the average number of times each buyer purchases the brand during the time period.  For example, if during the given time period, 1,600 purchase occasions are made by 500 buying households, Brand X’s purchase frequency is 3.2. 

Purchase size is the average product volume purchased each time the product is bought.  For example, if 3,000 total units were purchased on 1,600 separate occasions, Brand X’s purchase size is 1.9 units.

The major benefit that consumer panels provide is feedback on the direction of your marketing strategies.  For example, if you find that the penetration of your product is falling or is below competitive penetration rates, you can increase penetration with advertising, brand extensions, couponing, trial packs, product sampling and displays. 

A decline or deficiency in purchase frequency can be addressed with smaller package sizes, couponing, displays, in-pack/on-pack promotions, and sweepstakes/contests.  If you need to increase the purchase size, marketing tactics like two-for-one deals, bonus packs, trade deals, and larger packages will address this issue.

Behavioural Measures

In addition, panels offer the capability of measuring key behavioural activity that also influences sales.  These measures are deal propensity, repeat rate, purchase cycle and loyalty. 

Deal Propensity is the percentage of Brand X volume sold that a panellist reported, ‘on deal.’  Deals can represent manufacturer coupons, store coupons, store deals or other deals.  For example, if 5,000 units purchased were reported as ‘on deal’ out of a total volume sold of 20,000 units, then 25% of Brand X sales are ‘on deal.’ 

Repeat rate is the percentage of buyers that make 2 or more purchases during a given time period.  This is a very important measure when introducing a new product.  If Brand X has 10,500 total buyers and 3,500 of those buyers have purchased Brand X at least twice, the repeat rate for Brand X is 33%.

Purchase cycle among repeat buyers is another key behavioural measure.  It is often confused with purchase frequency.  However, purchase cycle only considers the number of purchase occasions among repeat buyers, while purchase frequency considers the number of purchase occasions among all buyers, including one-time buyers who have a frequency of 1.0.  Purchase cycle is expressed in number of days and gives a feel for the consumer usage rate of your product.

Brand loyalty is another important behavioural measure.  It represents the percentage of category volume purchases among item buyers who are satisfied by that item.  This important diagnostic measure is frequently included as a key business objective in growing brand volume.  Loyalty rates are calculated by dividing the brand volume purchased by the total category volume purchased among Brand X buyers.  For example, if a Brand X buyer purchases 10 units of Brand X out of a total category purchase of 20 units, that buyer is 50% loyal to Brand X. 

There are a number of additional factors – including category purchase frequency, time period purchased, number of competitive items and penetration strategies – that impact loyalty.  Low frequency categories tend to have high brand loyalty numbers because of the many one-time buyers.  Shorter time periods show higher loyalty because there is not enough time to observe switching behaviour.  Fewer competitive items generally lead to higher loyalty, as there is less opportunity for switching.  And differing penetration strategies may affect loyalty as some households take advantage of promotions and offers.  For these reasons, it is best to look at loyalty on an annual basis.

Household panel data provides an excellent complement to store-level scanning data.  Scanning data provides you with key measures of what happened to your business – sales up or down, share up or down.  Household panels, on the other hand, help complete the picture of the consumer’s behaviour – giving the “why behind the buy.”  Why did sales increase?  To which competitor did my brand lose share?  What caused my sales to decline – purchase frequency or purchase size?  Most important, household panel data provides a strong direction for manufacturers and retailers to develop their marketing strategies. 

Note: This article is based on ACNielsen’s Homescan consumer panel in the US, although the same principles and techniques apply in the UK.

For further information contact Kim Dunkason, ACNielsen (www.acnielsen.co.uk) on Tel. 01865 732275.
Email: kim.dunkason@acnielsen.co.uk

Date article published: 31/007/2002