|
Bear with me here, readers, I'm
about to throw a whole bunch of numbers at you!
All hell broke loose after my
article titled 'ARE
USTA AND TIA COUNTING PLAYERS THAT DON'T EXIST?' appeared last month.
If you recall, I took the
number of U.S. ball sales and determined no way do we have over
22 million players (adults and kids). 5-7 million seemed much
more likely to me.
And I pointed to that ominous
PAC study that seemed to predict way more tennis players than
are actually playing according to those ball sales.
Well, I was able to talk to a
representative of Sports Marketing Surveys in Jupiter, Florida.
I appreciate that he took the time explaining to me how those
surveys are conducted. (Thank you, Keith.) However, I wasn't
quite convinced. Back in my mind the thought that whoever pays
for such a study has a vital interest in keeping those numbers
high, did not go away.
In last month's FEATURE 1 I wrote,
"Participation numbers are fuzzy, at best, given the sampling
of only 20,000 people by the Physical Activity Council (PAC)."
That's only 20,000 people for all sports surveyed. Similarly,
when they survey European countries, they call a sampling size
of 3,000 per country "robust research" whereby the
location of those households should make a big difference. No?
Example: The area of Greater
Atlanta in Georgia has more tennis player than any other state
in the Union, more than Northern and Southern California combined,
and most likely more than a lot of countries in Europe and South
America. If you call 1,000 households in Atlanta you will get
a survey result that will be considerably different than one
in Los Angeles or New York.
But, I am not a statistician,
of course. I want to post a letter below from a person who knows
much more about this than me. And he is also a USPTA Master Professional/Business
and Economic Research Manager from Colorado. Make up your own
mind when you read Gary Horvath's analysis.
|
I applaud the
TIA for publishing its annual report on the state of the industry.
It has value for the industry, but it also has limitations.
First, there was an average annualized change in participation
of -0.4% from 2010 to 2018. Between 2010 and 2018 the population
increased by 17.8 million players, an annualized rate of growth
of .70%. This increase in population and decrease in tennis participation
caused "market share" for tennis to decline from 6.0%
in 2010 to 5.5% in 2018. Have the efforts of the industry prevented
the decline in participation from being worse or have they simply
been ineffective?
Second, the Tennis Equipment Index currently sits at 96. This
is below the value of 100 in 2003, when the index was started.
The TEI is expressed as the nominal value of sales, i.e. it does
not account for inflation. Between 2003 and 2018, U.S. inflation
increased at an annualized rate of 2.1%. This suggests the decrease
in sales is greater than shown. How are the manufacturers able
to provide quality equipment in this environment?
Third, participation in Cardio Tennis is irrelevant, despite
its connection to TIA. It would be more beneficial to know how
people are involved in tennis and other sports. Are they taking
private lessons, group lessons, playing leagues or tournaments,
or do they want to play in an unstructured setting? Do they really
care about the USTA and USTA programs? What do they like and
don't like about these activities? Should TIA review its survey
and collect additional information about what the customer expects
from their tennis experience?
Fourth, I have difficulty reconciling some of the equipment and
participation data with what I see on the streets and read in
other research. At the top of the list is the junior participation
numbers for 13-17 year old players. Most research shows that
adolescents start dropping out of sports at the age of 14. If
the TIA numbers are correct, then the growth in this market segment
bucks the trend. It would be helpful to hear the 13-17 year old
players describe why they are attracted to tennis. It seems presumptuous
to assume that this increase is driven by new or existing programs.
I also have trouble understanding the relationship between the
unit sales numbers and participation numbers. Does it make sense
that 17.2 million players are only purchasing 2.2 million racquets
and 2.8 million sets of strings annually? Said differently, does
it make sense that only 12% and 16% of the players are buying
new racquets or strings annually?
My final "data concern" deals with the participation
and sales data for 2008-2010, the time of the Great Recession
and the start of the recovery. How is it that tennis participation
and sales expanded during the downturn, but it has declined during
the recovery?
In closing, I applaud TIA for producing the report. I hope stakeholders
understand the message it delivers. Having said that, there are
pockets of growth within the U.S. and there are obviously many
areas at the other end of the spectrum. As if the message in
the report isn't enough, how is the tennis industry going to
address demographic changes in our country and the declining
fertility rate? These factors will have a significant impact
participation and sales in years to come.
The bottom line - Given the stagnancy/decline in participation
and sales, how are stakeholders going to increase sales and participation
in the industry over the next decade?
Gary Horvath |
In an earlier communication,
Gary Horvath had suggested that he is "confident the methodology of Sports
Marketing Research is sound, although given the number of categories
(117 sports, age groups, frequency of play, etc.) it is possible
the sample size may be too small."
Well, according to Keith, the
sample size for the PAC study is 20,000. Sounds big? Check this
out: they break all those people down into 9 U.S. regions, 6
or 7 age groups for men, 6 or 7 age groups for women, several
incomes per group and several ethnicities. Still call that big?
I call that voodoo math.
As far as the ball sales question
is concerned, Gary Horvath summarized it all up with a keenly
practical way of looking at data: "The 17.84 million players had an
average of 21.5 play occasions per player. (The latter is an
absolutely worthless statistic). The 383.9 million play occasions
help us get to the ball numbers. If all of those were doubles
occasions and a new can of balls was used for each play session
then 95,975,000 new cans of balls were used. On the other hand,
if each of those sessions was singles then 191,950 new cans of
balls were used. From here it is possible to build all kinds
of scenarios. Most likely the 108 million units seem low."
I'd say it was probably a mixture
of singles and doubles play occasions. Averaging both numbers
gets me to 143,737 cans of balls or 431,211 units. About 4 times
as much as reported ball sales. And I'm not even counting the
other ball sales for collages, tennis academies, etc.
Greg Mason, CEO of HEAD Penn Racquet Sports, a
man whose word I trust, has assured me the ball sales numbers
are correct, although it has to be increased slightly because
Walmart and other branded balls are not included. That was very
helpful, of course, because taking ball sales into the equation
as a constant, we can determine with high probability that it's
the participation numbers that are overstated. Maybe by a lot.
Unfortunately, meaningful analysis
of PAC/TIA tennis data is almost impossible. Just look at the
fuzzy interpretation of what constitutes frequent or occasional
players. Are non-playing spectators counted? Is someone who played
last 5 years ago, counted as an occasional player?
Since the USTA has a vital interest
in keeping those numbers high and is also for the most part paying
for the studies, a lot of people I speak with indicate that fuzziness
may be done by design.
GERMANY
I have been trying to get reliable data for tennis participation
in other countries. No one is willing to give you clear answers.
Maybe no one can? A DTB executive said Germany has 1.38 Mio.
club members, 1 Mio. in private facilities, 1.5 Mio. in holiday,
companies, 4 Mio.. more interested. No idea what to do with those
numbers. Can anyone tell me with a high degree of accuracy how
many people play tennis in Germany? If I take the average of
those first 3 numbers, I get to (maybe) 1.3 million players.
At a population of 82.79 million, the tennis population would
be 1.55%. Confidence in the accuracy of that number is quite
low. It's probably less.
-----------------------
GERMANY: MAX. 1.55% OF POPULATION PLAY TENNIS -------------------------
AUSTRALIA
Tennis Australia is all wishy-washy about their numbers. No reliable
data seems to be available. Is it 750,000 adult recreational
players or closer to a million? A trustworthy Australian tennis
professional tells me it is closer to 500,000 and maybe even
less. OK, I'm taking 600,000 as the most likely real number and
divide it by a population of 24.6 Mio. people. That makes it
2.4% of the Australian adult population plays tennis.
-----------------------
AUSTRALIA: 2.4% OF POPULATION PLAY TENNIS -------------------------
USA
The TIA reports 17.2 million adult players. At a population of
327 million people, the United States must have a 5.25% tennis
population. Does that sound realistic?
Keith mentioned that participation
rate across all sports is 6%. Sports Marketing Survey is taking
this number, applies it for tennis, and uses the data from a
survey sample that seems to be too small, to determine all numbers
that eventually lead to 17.2 million U.S. tennis players. Hmmmmhhh.
----------------------- U.S.A.:
5.25% OF POPULATION PLAY TENNIS? --------------------------
|
USTA League
Unique Players |
According to the USTA, 318,139
unique players were registered in USTA Adult Leagues last year.
Would you say those players should all represent CORE players?
Remember, according to the TIA, CORE players play 10 or more
times per year.
Consider this: What do you think
the percentage of USTA Adult League players is compared to the
rest of the CORE player population in any given area (except
in Atlanta)? Averaged throughout the United States. 50%? 40%?
30%? At the time I worked on the USTA Tennislink Team they said
the number is 10%. I really never trusted that number but let's
say it's 30%. Sounds realistic? Maybe lower in Arkansas, maybe
higher in California.
Now consider this: If 318,139
USTA Adult League players represent 30% of all CORE players,
the CORE total should be 1.06 million. Right? Well, the TIA published
number is 9.67 million. How do you explain that?
In other words, according to
the TIA, the 318,139 USTA Adult League players are only 3.29%
of all CORE players in the U.S.A. Who can possibly believe that,
folks?
|
Who is pulling
the wool over our eyes, folks? |
|
