Out of Home Advertising Effectiveness and Return of Investment

Out of Home Advertising Effectivenessand Return of Investment (ROI)

oaaaOutdoor Advertising Association of

America, Inc.

© 2011 Outdoor Advertising Association of America, Inc. 1

Out of Home Advertising Effectiveness and Return of Investment (ROI)

Global ROI Analysis Over the past three years, the British media research firm BrandScience analyzed over 600 econometric case studies to gauge the impact of out of home on sales results. These case studies represent national and local campaigns from Australia, Denmark, Finland, Germany, Norway, Sweden, UK, and the United States. The information provides an international trend analysis on ROI with quantitative insights demonstrating how different media drive sales independently and in a media mix. It is a statistical analysis process which identifies and measures the inputs that have caused something to change. The information is used to provide insights about how out of home advertising works relative to other media, independently and in synergy with other media. First, BrandScience identified and collected key data from the case studies. They then determined which factors are statistically significant and the magnitude of each factor’s impact on sales. After analyzing these case studies, BrandScience found out of home advertising has a high return on investment (ROI) across all categories. For information on econometrics, please see Appendix I on page 12. On a global scale, BrandScience found for each dollar spent on out of home advertising, an average of $2.80 was received in sales. Television and print advertising have a lower ROI, yet receive a greater share of the dollars in the average media mix.

$2.43 $2.41

$3.14 $3.38

$2.80

$-

$0.50

$1.00

$1.50

$2.00

$2.50

$3.00

$3.50

$4.00

TV Print Radio Online Outdoor

Sal

es R

OI

Global

41.4 23.3 12.8 9.1 13.5

Average % Media Mix


Out of home advertising has a particularly strong sales effect for telecommunications, media, retail, and travel advertisers. If production costs are taken into account, out of home advertising’s ROI would be even better compared to costly media, such as TV.

$3.89 $4.01 $3.52

$2.67

$4.72

$-

$1.00

$2.00

$3.00

$4.00

$5.00


Sal

es R

OI

Telecom

$0.60 $0.51 $0.54

$0.42

$1.40

$-$0.20 $0.40 $0.60 $0.80 $1.00 $1.20 $1.40 $1.60


Sal

es R

OI

Media

$3.89 $3.50

$2.80

$3.94 $3.79

$-

$1.00

$2.00

$3.00

$4.00

$5.00


Sal

es R

OI

Retail


BrandScience also found for each week after an out of home campaign, there is an average of 55 percent retained, beating print and radio.

US ROI Analysis Looking specifically at the United States, BrandScience analyzed 43 case studies featuring out of home advertising from the World Advertising Research Center (WARC) database, which features campaigns entered into awards programs for effectiveness. While not all of these campaigns were award winners, they were effective -- cases featuring out of home had a higher sales effect. BrandScience also found when there is a higher proportion of out of home in the mix, effectiveness increases. This may be due to two factors:

1. Out of home advertising is an effective reach medium. 2. When used in combination with other media, out of home not only extends the reach, but

also reinforces an advertising message.

$3.91

$2.95

$4.02

$3.01 $3.55

$-

$1.00

$2.00

$3.00

$4.00

$5.00


Sal

es R

OI

Travel

49%

35%

55%

Print Radio Outdoor

Ret

enti

on

Aft

er O

ne

Wee

k


According to BrandScience, the percentage increase in sales triples when campaign spending on out of home advertising moves from a low amount to a medium amount. Sales more than double when a high amount is spent on out of home. BrandScience found the optimal return on investment is gained when a campaign’s overall ad spend is low, but the proportion of out of home in the media mix is high.

Sales Increase by Out of Home Proportion of the Media Mix

Low Medium High

% I

ncr

ease

in S

ales

Campaign Dollars Spent on Out of Home

Sales Increase By Out of Home Usage In Media Mix

campaign spend low campaign spend medium

campaign spend high

% I

ncr

ease

in S

ales

Out of Home Proportion of the Media Mix

outdoor spend low outdoor spend medium outdoor spend high


Out of home advertising lowers the cost of an advertising program. The chart below shows the relative costs of one thousand impressions (CPM) among adults 18+ for major media.

Source: Wells Fargo Securities, 2007-2008

$2.10

$7.80

$5.00

$9.00

$13.00

$17.50

$10.30 $12.60

$14.70

$22.30

$9.50

$18.50

$25.00

$2.30 $3.40

$8.40 $10.90

$11.10

$21.30

$28.00

$43.00

$22.50

$56.50

$-

$10.00

$20.00

$30.00

$40.00

$50.00

$60.00

Radio

Online

Broadcast TV

Cable TV

Outdoor

Magazines

Newspapers

Direct MailCPM Comparisons of Major Media


Using out of home in combination with other media improves the decay rate, or time it takes for a person to forget a campaign’s advertising message. This works particularly well for TV and online advertising. When out of home advertising is included, the retention rate increases by several days.

Not only is the campaign effective for a longer period of time, but according to the 2010 study, How Out of Home Advertising Works, by MarketShare Partners and Clear Channel Outdoor, out of home advertising is 47 percent more effective when online carryover is taken into account. For example, a person sees a product advertised on a billboard, which leads them to research the product online, where the consumer is exposed to targeted online ads, and then purchases the product. While this may be attributed to Internet advertising, out of home’s indirect effect should be noted.

6

8

13

17

0

2

4

6

8

10

12

14

16

18

20

Without Out of Home With Out of Home

Mes

sag

e D

ecay

Rat

e (D

ays)

Internet TV


Individual Campaign ROI Analysis Overall, out of home media has shown to be effective in producing consistent ROI results compared to other media, both globally and nationally. Similar results are evident when considering specific brands. BrandScience conducted a detailed ROI analysis of a prominent hotel chain. Hotel Chain X was a successful brand with an expanding portfolio of properties which

advertised across a range of media. At the end of 2005, Hotel Chain X saw that the hotel industry as a whole had enjoyed

significant growth since 2003 and Hotel Chain X’s sales performance had been comparable to the overall market.

In view of this, there was a perceived need to assess the effectiveness of the marketing activity Hotel Chain X had undertaken to ensure that it had made a contribution to sales.

Three key performance indicators were chosen and statistical models were built to estimate the effects of marketing activity:

– Occupancy – the % of hotel rooms in use – ADR – the average daily rate charged for hotel rooms

To quantify the impact of marketing, statistical models were built to estimate the effect on the individual performance indicators of different marketing activities. These models work by relating the change in an influencing factor – such as television advertising, out of home advertising, or improvements to the hotels themselves - to the change in the measured indicator, while taking into account all other influencing factors. See Appendix I on page 12 for a detailed explanation of the modeling.


The models analyzed multiple variables to look at the relationships over time as opposed to cross-sectional, which would typically look at the relationship across markets. All effectiveness findings/ROI figures developed were estimated from model coefficients that were statistically significant at a 95 percent confidence level. Over the period, Hotel Chain X improved its occupancy rate significantly over its competitors, although those competitors were engaged in the same types of activity as Hotel Chain X during this period. The occupancy model demonstrates the extent to which marketing activity enabled Hotel Chain X’s performance in this competitive environment.

65%

67%

69%

71%

73%

75%

Occ

upan

cy

Hotel Chain X Vs Competitors Occupancy Rates

Hotel Chain X Underlying trend in OccupancyCompetitors Underlying Trend in Occupancy


The base level in grey represents trends, seasonal effects and occupancy that would happen without pricing, or marketing fluctuations. This is the occupancy level that would have been achieved without any marketing activity. The green line at the bottom represents the negative effect of competitor activity. The colored lines at the top of the chart show the effects of controllable variables – marketing, pricing, distribution, hotel changes, etc. Within these trends, the pink line at the top shows the contribution out of home advertising made to Hotel Chain X’s occupancy levels.


The contribution made by different media to occupancy levels varies by media format. Compared to the other media used, the impact of out of home, represented by the thick, orange area at the top of the graph, is large.

Marketing was estimated to have driven an increase in occupancy of nearly 10 percent with a quarter contributed by out of home advertising. Media Spend Media Effect

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

12-J

an-0

3

12-F

eb-0

3

12-M

ar-0

3

12-A

pr-0

3

12-M

ay-0

3

12-J

un-0

3

12-J

ul-0

3

12-A

ug-0

3

12-S

ep-0

3

12-O

ct-0

3

12-N

ov-0

3

12-D

ec-0

3

12-J

an-0

4

12-F

eb-0

4

12-M

ar-0

4

12-A

pr-0

4

12-M

ay-0

4

12-J

un-0

4

12-J

ul-0

4

12-A

ug-0

4

12-S

ep-0

4

12-O

ct-0

4

12-N

ov-0

4

12-D

ec-0

4

12-J

an-0

5

12-F

eb-0

5

12-M

ar-0

5

12-A

pr-0

5

12-M

ay-0

5

12-J

un-0

5

12-J

ul-0

5

12-A

ug-0

5

12-S

ep-0

5

12-O

ct-0

5

Direct Mail (Stays) Internet Spend @ 40% Carryover

Newspaper Spend @ 80% Carryover Mags Spend @ 80% Carryover

Outdoor Spend @ 80% Carryover

14.6

20.8

3

7.9

3.9

0.3 2.1 Out of Home

TV

Newspapers

Magazines

Internet

Radio

Direct Mail

2.4 2.7

0.7

1.61.40.02

0.8


As might be expected, marketing has less of an effect on room price (ADR) than on occupancy levels. The estimated total contribution was less than 2 percent with over half attributed to out of home. If increasing room price – in effect persuading people to trade up, or buy at peak times, is the goal – out of home and Internet should be increased at the expense of magazines, newspapers and radio.

Spend ($M) % Contribution ADR Effectiveness index

Out of Home 14.6 0.9 202

TV 20.8 0.1 160

Newspapers 3 0 0

Magazines 7.9 0.2 83

Internet 3.9 0.3 252

Radio 0.3 - 55

Direct Mail 2.1 0.1 156

Total 52.46 1.6 100


Appendix I

Econometrics involves evaluating data on all potential drivers of sales. Once the contributing factors are identified, they are assigned a weight to determine the amount each factor contributes. Once the mix is known, past volume is explainable and future volume is predictable

For example, in trying to explain volume sales of a brand, why does the line go up some weeks vs. other weeks?

-1500

-1000

-500

0

500

1000

1500

2000

2500

3000

3500

2000

Jul

2000

Sep

2000

Nov

2001

Jan

2001

Mar

2001

May

2001

Jul

2001

Sep

2001

Nov

2002

Jan

2002

Mar

2002

May

2002

Jul

2002

Sep

2002

Nov

2003

Jan

2003

Mar

2003

May

2003

Jul

2003

Sep

2003

Nov

2004

Jan

2004

Mar

2004

May

KPI


-1500

-1000

-500

0

500

1000

1500

2000

2500

3000

3500

2000

Jul

2000

Sep

2000

Nov

200

1 Ja

n

2001

Mar

2001

May

2001

Jul

2001

Sep

2001

Nov

200

2 Ja

n

2002

Mar

2002

May

2002

Jul

2002

Sep

2002

Nov

200

3 Ja

n

2003

Mar

2003

May

2003

Jul

2003

Sep

2003

Nov

200

4 Ja

n

2004

Mar

2004

May

Residual KPI Model

R2 = .01728

-1500

-1000

-500

0

500

1000

1500

2000

2500

3000

3500

2000

Jul

2000

Sep

2000

Nov

200

1 Ja

n

2001

Mar

2001

May

2001

Jul

2001

Sep

2001

Nov

200

2 Ja

n

2002

Mar

2002

May

2002

Jul

2002

Sep

2002

Nov

200

3 Ja

n

2003

Mar

2003

May

2003

Jul

2003

Sep

2003

Nov

200

4 Ja

n

2004

Mar

2004

May

Residual KPI Model

R2 = .37884

-1500

-1000

-500

0

500

1000

1500

2000

2500

3000

3500

200

0 Ju

l

2000

Sep

200

0 N

ov

2001

Jan

2001

Mar

2001

May

200

1 Ju

l

2001

Sep

200

1 N

ov

2002

Jan

2002

Mar

2002

May

200

2 Ju

l

2002

Sep

200

2 N

ov

2003

Jan

2003

Mar

2003

May

200

3 Ju

l

2003

Sep

200

3 N

ov

2004

Jan

2004

Mar

2004

May

Residual KPI Model

R2 = .82771

-1500

-1000

-500

0

500

1000

1500

2000

2500

3000

3500

2000

Jul

2000

Sep

2000

Nov

200

1 Ja

n

2001

Mar

2001

May

2001

Jul

2001

Sep

2001

Nov

200

2 Ja

n

2002

Mar

2002

May

2002

Jul

2002

Sep

2002

Nov

200

3 Ja

n

2003

Mar

2003

May

2003

Jul

2003

Sep

2003

Nov

200

4 Ja

n

2004

Mar

2004

May

Residual KPI Model

R2 = .92820

First, seasonal events are accounted for. Then, print media, out of home advertising, TV, and other factors.

Seasonal events account for a portion of sales Accounting for print advertising brings the fluctuations. model closer to actual sales.

Adding out of home advertising to the model With TV, the model is almost complete. creates an even better fit.


Finally, other factors are taken into account.

The equation can now be used to break out sales by contributing factors.

-1500

-1000

-500

0

500

1000

1500

2000

2500

3000

3500

2000

Jul

2000

Sep

2000

Nov

2001

Jan

2001

Mar

2001

May

2001

Jul

2001

Sep

2001

Nov

2002

Jan

2002

Mar

2002

May

2002

Jul

2002

Sep

2002

Nov

2003

Jan

2003

Mar

2003

May

2003

Jul

2003

Sep

2003

Nov

2004

Jan

2004

Mar

2004

May

Residual KPI Model

R2 = .94478

-2,500

-1,500

-500

500

1,500

2,500

3,500

Jul-0

0

Sep

-00

Nov

-00

Jan-

01

Mar

-01

May

-01

Jul-0

1

Sep

-01

Nov

-01

Jan-

02

Mar

-02

May

-02

Jul-0

2

Sep

-02

Nov

-02

Jan-

03

Mar

-03

May

-03

Jul-0

3

Sep

-03

Nov

-03

Jan-

04

Mar

-04

May

-04

Base, Seasonality + other factors DMPress OutdoorTV Competitor 1Competitor 2


Appendix II

BrandScience used the WARC (World Advertising Research Center) database case study section for the US study. OMG has a subscription to this database and can precisely analyze information, but not circulate any direct content. BrandScience filtered the WARC database to case studies in the US that contained a reference to billboard, poster or out of home. These case studies, generally around 1,000 words in length, were then examined to verify that the references referred to the case itself rather than to comparative cases, and to see what data they contained. There were more than 100 case studies in the original list for review, but BrandScience narrowed the final list to around 50 case studies which were right in terms of geography, featured out of home, and also crucially provided some hard metrics on both spend and results. However, one of these proved to be a duplicate of the results from part of another case study (Altoids) and six others gave results which BrandScience could not factor into the final model, generally because they were attitudinal measures, not sales measures. They were Allstate Corp., Budweiser, FCUK, Holiday Inn, National Thoroughbred Racing Association and Volkswagen. BrandScience recognizes that by selection of the WARC database the study was initiated with a basis of campaigns likely to be high achievers (though achievers in general, not because of them medium). The cases were not all award winners, but award entrants. However, the database provides hard information on spend and result. BrandScience lacks a “base” figure, actually two base figures:

- The average sales effect of a campaign not using out of home, in the WARC database of case studies. - The average sales effect of a campaign, not an award entrant. To adjust for this factor, an academic meta-analysis by Sethuraman and Tellis (1991) was cited which found the average advertising elasticity to be 0.109


BrandScience analyzed the following brands and campaigns in the United States from data in the WARC database. The campaigns span a 30 year time period, and all feature out of home at varying levels.

Apple – Silhouette Apple – Think Different Lego – Bionicles California Milk – Happy Cows Altoids – Curiously Strong Carlo Rossi – Jug Simple Coca-Cola – Beautiful People Coca-Cola – Zero Taste Similarity DHL – Competition. Bad for Them. Great for You. ESPN – Monday Night Football: Is it Monday yet? Gap – For Every Generation Gap – Khakis H&R Block – Worried About Bill Heineken – It’s All About the Beer Ikea – Unböring Infiniti G35 – All Wheel Drive Launch Holiday Inn – Rejuvination Las Vegas Convention & Visitors Authority – Vegas Stories Campaign Levis – 501 Reasons Levis Dockers – Nice Pants McDonald’s – I’m Lovin’ It Mercedes Benz – Unlike Any Other Mercedes Benz – Corporate Branding MINI Cooper – Zug. The Other MINI Mitsubishi – Wake Up and Drive Montana Meth Project – Not Even Once Napster – It’s Coming Back Nissan – Enjoy the Ride Palace Sports & Entertainment: Detroit Pistons – Going’ to Work. Every Night Partnership for a Healthy Mississippi – Question It Peace Corps – Life is Calling. How Far Will You Go? PepsiCo – The Not-So-Vanilla Vanilla Rock the Vote – Yes/No Ballot Box Saab – Born from Jets Schick-Wilkinson Sword – Shaving Made Simple Master foods – Snickers Marathon. The Energy You Crave Kellogg’s – Special K Kick-Star Diet Plan Sprint Nextel – Done Tourism New Zealand – 100% Pure New Zealand Toyota – Who Could Ask for Anything More TV Guide – On the Inside Wachovia – Uncommon Wisdom Wrigley’s Orbit – No Matter What

Documents

Out of Home Advertising Effectiveness and Return of Investment