Group assignment for Marketing Research class at KU Leuven.
We applied trendspotting method to test possible explanations of beer market decline in UK.
4. ‘The UK population drinks less since 2004’:
DRINKING IN
PUBS BECAME
TOO
EXPENSIVE
LESS
SUITABLE
FOR
IN-HOME
DRINKING
LACK OF
INNOVATION
Source: Mintel report, ‘Beer – UK’, December 2011
6. HYPOTHESIS
1. Market declines because of
economical crisis
2. Market declines because of healthy
lifestyle of consumers
3. Market declines because of bad
weather
*assumption: if people google ‘beer’, they will buy it
10. H1: People search for „cheap beer” in times of crisis
Cheap beer
Cheap beer
Beer offers
Tesco beer Beer prices
Beer deals
Crisis
GDP
Unemploy-
ment rate
11. 1) Google Trends: search for terms related with cheap beer
2) Google Public Data Explorer: indicators for crisis
3) Check correlations between crisis and cheap beer indicators
H1: People search for „cheap beer” in
times of crisis
12. • When unemployment rate gets higher, more people search
for beer offers (r=0,91)
• When unemployment rate gets higher, less people search for
beer prices (r=-0,63)
• GDP is positively correlated with Google users interest in
„cheap beer” (r=0,43)
H1: People search for „cheap beer” in
times of crisis
cheap beer beer offers tesco beer beer deals beer prices
Unemployment
rate
-0,20454 0,909714 0,464214 - -0,62532
GDP 0,432173 -0,23236 0,146774 - 0,271101
13. H2: People are not interested in buying
beer because they want to live healthy
„Healthy
lifestyle”
„Beer”
14. 1) Google Trends: search for „heathy lifestyle”
2) Google Trends: search for term „beer”
3) Check correlations between beer and healthy lifestyle
H2: People are not interested in buying
beer because they want to live healthy
r = -0,23
15. H3: People drink less beer because of bad weather
Weather
Min.
temperature
Rain (mm)
Max.
temperature
Sun hours
„Beer”
16. 1) Google Trends: search for „beer”
2) Meteorological office UK: temperature, rain, sun (Oxford)
3) Select meteorological data only for summers (June – September)
4) Calculate average max. and min. temperature, sum of rain mm, sum
of sun hours for each summer (2006-2012)
5) Calculate average number of „beer” searches in each summer
6) Calculate correlations between „beer” searches and weather
indicators
H3: People drink less beer because of
bad weather in summer
17. Max. temp Min. temp Rain (mm) Sun hours
„Beer” 0,069662 0,075526 0,354535 0,299066
Summer (OXFORD) "Beer" Max temp Min temp Rain mm Sun hours
Jun - Sep 2006 83 23,525 13,3 242,1 873,6
Jun - Sep 2007 74 20,5 11,45 240,2 695,9
Jun - Sep 2008 70 20,175 12 307,8 564,1
Jun - Sep 2009 69 21,825 11,925 177,7 769,9
Jun - Sep 2010 72 21,775 11,925 251,1 676,3
Jun - Sep 2011 81 20,525 10,625 194,6 677,4
Jun - Sep 2012 84 19,9 11,75 394,9 648,1
1. Averages and sums for beer search and weather indicators
2. Correlations between „beer” and weather indicators
18. WEAKNESSES OF THE METHOD
• Hard to apply for FMCG
• Different data sources = different time span
• Noise – people search for beer also when they
mean i.e. beer festivals, beer pong, beer
glasses
• Data exported to excel file inconsistent with
graphs presented by Google Trends
• Relative numbers instead of real number of
searches