CONTENTS

Evaluating PromoFlor: Has Promotion Had an Impact?

Manager's Viewpoint

The Rest of the PromoFlor Story...

Editor's Notes

Director’s Corner

Next Meeting

Evaluating PromoFlor: Has Promotion Had an Impact? *

by Ron W. Ward

In late 1993, legislation was approved for the implementation of a national program to generically promote fresh cut flowers and greens in the U.S. domestic market. All programs were underwritten through an assessment of .5 percent on qualified handlers with annual sales of $750,000 or more. The first commercials, aired on national television in late February 1996, introduced “BuzzTM” and the “Think FlowersTM” campaign. These programs were administered through the National PromoFlor Council.

Generic promotion of fresh flowers is unique among the current national checkoff programs because it is the first program designed solely to promote the aesthetic value of a product. For many commodity checkoff programs, the promotion message is directed primarily to informing potential consumers about the physical attributes of the product. Promoting flowers is intended to change perceptions, awareness, and recall with a message focusing on the pleasures received from buying and giving flowers.

Understanding Demand
Has the promotion of fresh cut flowers and greens changed the demand for flowers and greens? To answer this question, we need to have a clear understanding of demand and how to measure promotion’s impact on it. While flowers are purchased throughout the year, there are clearly special occasions such as Valentine’s and Mother’s days, etc., where purchasing activities increase several fold. We refer to these buying patterns as seasonal factors in demand reflecting consumer habit persistence. The quantity purchased also depends on price, consumer income, inflation, and other economic and demographic factors that cannot be easily changed, especially by the flower industry. This group is referred to as the external factors of demand. Finally, one may attempt to change perceptions, awareness, and recall through promotions, product development, and innovations. We refer to this third group as the controlled factors of demand. PromoFlor fits into this third group. Demand is then defined as the relationship between the purchases of flowers and the three categories of factors influencing demand; to measure it, the empirical linkage between consumers and the seasonal, external, and controlled factors must be known. For PromoFlor specifically, one must measure how much of a change in demand can be attributed to the generic promotion of fresh cut flowers while recognizing that other factors also cause changes in demand.

Total demand for flowers can change in basically two ways:
(a) The number of buyers changes from period to period with the entry of new buyers, and increases in the number of buyers with a history of purchasing flowers. The frequency of buying is expressed as a percent of U.S. households buying flowers within a given time period. Likewise, the numbers can be shown across demographic groups such as income categories.
(b) The quantity and expenditure per buyer reflect the intensity of consumption. Demand rises and falls with changes in the purchases per buyer.

To measure the potential effects of PromoFlor, it is essential to determine if the generic promotion of fresh cut flowers attracted additional buyers to market (a) above, and if the advertising generated additional purchases per buyer (b) above.

How is Demand Measured?
To measure demand, data on consumer purchasing activities must be available. Given the data, statistical tools, specifically, econometrics, must be used to determine the linkage between purchases and promotion. In our case, models depicting the relationship between demand for flowers and generic advertising are estimated. Empirical models are then used to calculate a rate-of-return to the programs.

Potential buyers of fresh cut flowers in U.S. households are grouped into four income categories: (1) under $25,000; (2) $25,000 to $49,999; (3) $50,000 to $74,999; and (4) $75,000 and over. Flower purchases are recorded for each income group along with prices, purchase period, and other economic factors. Using statistical models, the impact of PromoFlor advertising and promotion on flower sales (including the number of buyers and the amount of expenditures within each income group) is estimated.

Since assessments occur at the handler level, it is essential for gains to be reflected at that level. To convert retail sales back to the handler level, we must determine the value of fresh cut flowers in retail sales of arrangements. It is assumed that 65 percent of the retail arrangement value is in fresh flowers (includes costs of the flowers and other costs associated with the flower component of the arrangement, and profits on flowers included in the arrangement). Whereas, for other flower sales, almost all of the product is in fresh flowers. Once the retail flower value from arrangements and other fresh cut flower sales is known, then the retail flower value is divided by an assumed industry retail markup to reflect the fresh cut flower value to handlers (i.e., the equivalent gain in handler revenues).

Using these procedures and the econometric models, a change in PromoFlor promotion leads to potential gains in retail markets. These gains are then expressed in terms of equivalent values to handlers. Handler gains from promotion are compared to the program costs, thus giving a measure of the rate-of-return to the programs.

Flower Purchasing Data
Data collected by the National Panel Dairy Group (NPD) and sold through the American Floral Endowment provide a rich source of information about the retail sales and number of flower buyers. NPD data are based on a consumer panel of several thousand households who report their purchases of flowers, including zero purchases. While the NPD panel is a representative sample of U.S. households, the panel does not cover all consumers. Hence, NPD’s national projections still represent only part of all domestic flower purchases. Our calculated impact for PromoFlor is based on these data and may in some cases understate the full impact, since other types of consumers including commercial purchases are not captured in the NPD panel.

Flower Types and Buying Patterns
For evaluation purposes, retail flower sales are divided into three categories: fresh cut flowers, flowering pot plants, and dry/artificial flowers. While promotion activities and assessments apply only to fresh cut flowers, promotion may influence the sales of all three types of flowers. In this evaluation analysis, we are dealing only with the impact of PromoFlor on fresh cut flower sales.

Fresh cut flowers accounted for 52 percent of retail flower sales. Within the fresh cut sales type, arrangements accounted for 55 percent of the total; single stems contributed 16 percent; bunches totaled 23 percent; and the remaining was 6 percent. When viewing fresh cut sales across our four income groups, the lowest income group of households accounted for 22.7 percent of the sales; 35.8 percent were from households with incomes between $25-$49,999; households between $50-$74,999 contributed 23.4 percent; and the highest income group equaled 18.1 percent for fresh cut flower sales.

Buyer patterns also differed by income groups. During an average month, approximately 3 percent of households in the lowest income group bought fresh cut flowers. For the $25-$49,999 and $50-$74,999 groups, the percent buying within a month averaged 4.3 and 6.2 percent, respectively. Finally, around 9 percent of households in the highest income group bought fresh cut flowers at least once during a month.

Comparing Buyer Behavior Across Years
While the percent of households buying in any one month was around 5 percent, on an annual basis the percentages were larger. In 1993, almost 23 percent of the households purchased some form of fresh cut flowers within the year. From 1993 through 1995, the percentage trended downward ending with 20.5 percent in 1995. During the full year of 1996, the downward trend reversed direction with 22 percent of U.S. households buying fresh cut flowers. An important issue is whether or not PromoFlor contributed to this positive change.

Around 7.7 percent of the households bought flowers on special occasions such as Mother’s day, etc. There have been only small changes in this percentage over the last four years. In contrast, changes in buying patterns during non-calendar occasions were more pronounced. Almost 20 percent of U.S. households bought flowers in non-calendar periods in 1993. This value declined to 16.8 percent in 1995 but increased to 18.2 percent in 1996. It should be noted that most of the gains in buyers in 1996 were for non-calendar occasions.

But what about the frequency of purchases per buyer? In 1996, on average, there were 3.3 purchasing occasions per buyer. This was an increase over the 1995 level of 3.16 occasions per buyer and about 3.5 percent above the 1993 level. While the rate of change was relatively small, it points to an improvement in 1996 over the period seasons. Again, the main question is if PromoFlor contributed to the gains seen in 1996.

The Economic Impact of PromoFlor
Using NPD data and the statistical models noted above, the effect of PromoFlor advertising on the demand for fresh cut flowers was estimated. The impact was seen in both attracting additional buyers to the retail market and in additional expenditures per buyer. These advertising effects differed across the four income groups. The statistical analysis showed fresh cut flower advertising and promotions had a positive and statistically significant influence on demand for flowers. Household expenditures on fresh cut flowers increased at the retail level. Using the procedures described earlier, these increases in retail sales can be expressed in terms of dollars of additional gross sales to handlers. These gross sales are the revenues generated back to the handlers before subtracting out handler costs and profits.

Additional retail revenues attributed to PromoFlor were expressed in terms of gains in handler revenues. Dividing these revenue gains to the handlers by PromoFlor’s expenditures on advertising and promotions gives a total of $6.62 (gross) in handler sales for each dollar of PromoFlor expenditure. That is, for each dollar of PromoFlor advertising and promotion expenditures, the models show $6.62 of additional handler gross revenues generated per promotion dollar (including all handler costs and profits). A net rate-of-return in terms of revenue occurs by subtracting the $1 promotion cost from the $6.62 gained per promotion dollar. Therefore, for the 1996 season (through November 1996), each dollar of generic promotion of fresh cut flowers generated $5.62 in additional handler revenues per promotion dollar.

This gross gain of $6.62 per promotion dollar includes both the costs of goods and handler profits. While costs may vary across handlers, an industry standard markup of 1.43 can be used to express handler revenue gains in terms of changes in gross profits. If the .43 represents gross profits in the handlers’ markup, then 30 percent of the handler sales are gross profits. Multiplying this percent times the $6.62 in revenue gained per promotion dollar gives an increase in handler gross profits per promotion dollar by PromoFlor. Using this percent, a dollar of PromoFlor promotions produced $1.98 (i.e., $6.62 x .30) in additional gross profits per promotion dollar before subtracting out promotion costs. The net gain in profits per promotion dollar was $.98 after subtracting costs.

To reemphasize, these gains are based on a portion of the total industry captured by the NPD data. For example, commercial sales are not included in the calculations. If PromoFlor has positively impacted sectors not in the NPD data, the rate-of-return is understated.

PromoFlor Gains Across Income Groups
Estimated gains differ across the four income groups as noted earlier. Having a clear understanding of where the gains were generated is essential to program design and revisions in subsequent periods. Around 31 percent of the total sales gains attributed to PromoFlor occurred among those households with incomes of less than $25,000 annually. The $25-$49,999 group accounted for almost 40 percent, while the next group was near 30 percent. These three groups accounted for 93 percent of U.S. households. Interestingly, the analysis showed there were no gains estimated among those households with incomes of $75,000 and greater. That is, this income group did not show a consumption response to the promotions-- one can only conjecture why.

A second dimension to understanding gains relates to changes in the number of buyers vs. the intensity of buying as discussed with points (a) and (b) earlier. Within each income group, gains can occur with new buyers entering the market and with additional expenditures per buyer. Knowing how promotion works in this context is extremely useful for longer term programming.

But what share of total gains can be attributed to increasing the number of buyers in each income category? First, the percent of buyers is not relevant for the highest income group since responses were not statistically significant. However, for the lowest income group, the analysis showed that 70 percent of the gains among those households resulted from additional buyers entering the market. The majority of gains in this income group resulted from more buyers, less in terms of the amount of purchases per buyer. For the middle two income groups, between 52 and 57 percent of the promotion gains were due to more buyers. More households were attracted to the market and the intensity of consumption per buyer also increased.

Gains in Fresh Cut Flower Buyers
The analysis clearly shows additional buyers have been attracted to the fresh cut flower market with the introduction of the national flower advertising and promotion campaigns. During a typical month, around 5 percent of U.S. households bought fresh cut flowers (percentage differs by income group as stated earlier). Using statistical models, the additional number of buyers attracted to the market in a typical month can be estimated.

On average, there were 4.5 million households per month buying fresh cut flowers. An estimated 460,000 additional households bought flowers. These additional buyers represent an approximate 10 percent increase in the number of buying households within the NPD database for a month. This gain is directly attributed to “Buzz” and the “Think Flowers” promotions. Among these additional buying households, nearly 40 percent were within the lowest income category with roughly 70 percent of the total additional buyers among the two lower income categories.

Conclusions
We started with the question, “Has PromoFlor impacted the demand for fresh cut flowers?” Given the statistical analysis based on the NPD database, the answer is yes. Specifically, at this juncture in the evaluation process, the following major conclusions are noted:

1. Statistically, PromoFlor has positively and significantly affected fresh cut flower demand, specifically among households with incomes under $75,000.
2. An estimated rate-of-return indicates $6.6 of additional handler revenues were generated per dollar of PromoFlor advertising and promotion expenditures. Subtracting out the cost of PromoFlor, a net gain of $5.6 in additional revenues per advertising dollar was realized at the handler level.
3. The gains attributed to PromoFlor’s programs resulted from both attracting additional buyers and increases in the expenditures per buyer. For the lowest income group, nearly 70 percent of the gain was due to additional buyers entering the market. The share was nearly 50 percent for the next two income categories.
4. PromoFlor’s advertising and promotions produced an estimated 10 percent increase in the number of households buying fresh cut flowers in a typical month. Over 40 percent of these additional buyers were within the lowest income group.

At this point in the evaluation, one must recognize that the total promotion experience was relatively short since PromoFlor only had one year of promotion activities to study. Statistically, the analysis indicated the “Buzz” and “Think Flowers” campaigns stimulated a response among the buying public.

[ top ]

* Footnote: A handler referendum on the PromoFlor program was initially scheduled for November 1997 but was later moved up to June 1997. All evaluation materials and other aspects of PromoFlor’s efforts were distributed to handlers. In June 1997, the industry voted through referendum to terminate PromoFlor, with a vote of 58 percent against the programs. Although the industry benefitted economically from the programs, there were equity concerns about sharing costs. As the legislation was written, handlers paid all assessments (particularly handlers with sales above $750,000). Yet it is clear that some of the promotion benefits were accruing at levels in the vertical distribution system (i.e., retail) not subject to the assessments. While we may never fully understand all the factors influencing the final vote, assessment equity was definitely a concern.

Ronald W. Ward is a professor in the Food and Resource Economics Department, University of Florida, Gainesville, Florida. The American Floral Endowment was instrumental in providing the data essential to this evaluation.