1IN DIGITAL, YOU NEED TO BEPERSUASIVE TO BE IMPACTFUL(MOST OF THE TIME)Millward Brown Digital April 2013
SUMMARY OF FINDINGSWe analysed US data from the Link for TV, print and digital databasesto understand how Link copy testing metrics relate to one another.For print and digital ads, all measures are more highly correlated witheach other, compared with correlations seen across measures for TV.It is clear that there are many similarities between these media in theUS:• Branding will always be strongly related to impact• Ads that are easier to understand tend to be better branded• New information is strongly linked to persuasionHowever, there are also some clear differences:• For print and digital ads, a persuasive message is one route toimpactful advertising. For TV, impact and persuasion are unrelated.• For print and digital ads, virality is related to both persuasion andimpact; whereas for TV, the relationship is much weaker
3THERE IS ALMOST NO RELATIONSHIP BETWEEN IMPACT ANDPERSUASION FOR TV ADS, BUT A SIGNIFICANT ONE FOR PRINT ANDDIGITAL ADS05101520253035400 10 20 30 40 50 60AwarenessindexPersuasion top boxTVBased on 5858 USA-English online ads r=0.10808590951001051101151201251300 10 20 30 40 50ImpactscorePersuasion top boxPRINTBased on 578 USA- English online ads r=0.54607080901001101201300 10 20 30 40 50ImpactPersuasion top boxDIGITALBased on 837 USA-English online ads r=0.47
4FOR PRINT AND DIGITAL, A PERSUASIVEMESSAGE IS ONE ROUTE TO HIGH IMPACT…….…if you‟ve got a persuasive message, focus on getting themessage across clearly, rather than extra creativity.Correlation betweenImpact/AI and PersuasionTV Print Digital0.10 0.54 0.47
5THE DEGREE OF INTER CORRELATION ACROSSLINK MEASURES BY MEDIAGiven the relationship between persuasion and impact, it is notsurprising to find Digital Link shows the highest degree of intercorrelation between measures; this level is also fairly high for Print.0.430.550.62TVPrintDigitalAverage degree of inter correlation
6Enjoyment and stopping power are key to impact for Digital ads andenjoyment is important for TV, whereas stopping power and differenceare important for PrintImpact diagnosticsCorrelation withImpact/AITV Print DigitalBranding* 0.64 0.81 0.90Enjoyment* 0.50 N/A 0.79Involvement* 0.44 0.58 0.38Understanding 0.23 N/A 0.71Stopping power* N/A 0.74 0.85Different to other ads 0.40 0.70 N/AIMPACTNote*: these relationships are dictated largely by the fact that these metrics contribute to our Awareness Index/ Impactsummary metricsBRANDING IS ALWAYS IMPORTANT FOR IMPACT
7A PERSUASIVE MESSAGE IS ONE ROUTE TOIMPACT FOR PRINT AND DIGITAL ADSRelevance, believability, difference and virality contributestrongly to impact for print and digital adsPersuasion diagnosticsCorrelation withImpact/AITV Print DigitalNew information -0.03 0.51 0.50Relevant 0.14 0.71 0.68Believable 0.25 0.72 0.78Brand really different 0.30 0.78 0.80Brand Appeal 0.17 0.74 0.72Tell other people 0.44 0.78 0.76Saw online would send on 0.41 N/A 0.67IMPACT
8NEW INFORMATION DRIVES PERSUASION STRONGLYACROSS ALL MEDIA. PERSUASION MORE LIKELY TOIMPACT VIRALITY FOR DIGITAL ADS.Brand appeal and Persuasion are similar measures for allmedia and correlate highly across each mediumPersuasion diagnosticsCorrelation withPersuasionTV Print DigitalNew information 0.70 0.72 0.62Relevant 0.63 0.60 0.65Believable 0.45 0.43 0.47Brand really different 0.68 0.64 0.46Brand Appeal 0.77 0.77 0.74Tell other people 0.41 0.60 0.68Saw online would send on 0.31 N/A 0.70PERSUASION
9FOR DIGITAL, SOME MEASURES REGARDED ASIMPACT DIAGNOSTICS ARE RELATED TOPERSUASION ….…notably enjoyment and stopping powerImpact diagnosticsCorrelation withPersuasionTV Print DigitalBranding 0.23 0.30 0.26Enjoyment 0.29 N/A 0.60Involvement -0.01 0.34 0.26Understanding 0.33 N/A 0.37Stopping power N/A 0.45 0.64Different to other ads 0.05 0.29 N/APERSUASION
APPENDIX - THE DATAAll data included here is based on USA – English online ads, and comes from thefollowing databases:- Link for TV, Link for Print and Link for DigitalData used to look at correlations between measures has varying base sizesdepending on the measures, but total data sets include:- TV 8389 ads (2001-2012)- Print 1528 ads (2000-2012)- Digital 914 ads (2005-2012)10
IMPORTANT NOTES……..This analysis is designed primarily to help us understand the Link researchsolution better, and the dynamics of different media on these measures.This is NOT a validation or a hierarchy of importance.We have NOT correlated Link measures against external success criteria (in-market performance or sales).Rather, what we have done is look internally to understand the dynamics ofour own metrics better.Also, although correlations tell us where relationships exist, they do not tellus which measure “drives” the other. For instance we tend to see quite highcorrelations between branding and understanding. This is helpful, but wecan‟t say whether ads that are easier to understand tend to be better branded,or whether clearer branding makes ads easier to understand.Data is based on US only and could vary by country.
WHAT THIS MEANS FOR HOW WE SHOULD ANALYSE LINKRESULTS ACROSS DIFFERENT MEDIA……..Based on our strong TV testing heritage we tend to make certain assumptionsabout how Link tests “work” and which measures we should use to supportand diagnose which other metric.This analysis makes it clear that there are variations across media so we needto analyse pre test results for each media with a different mindset, and weneed to be prepared to take a different approach to our story-telling.• For print and digital ads , if the ad has done well it is more likely to havedone well on most measures.• For print and digital, a persuasive message is one route to get impact (butthere are others)• For TV ads we are more likely to end up with more nuanced stories whereImpact and Persuasion are not inter-related.We know from CrossMedia that the overall impact for print and digital adsalone are unlikely to be as powerful as that of a TV ad, which currently hasthe ability to reach the biggest audience in one single showing.So TV advertising, as part of a campaign, can helpfully „prime‟ the consumerfor other media to reinforce and enhance the brand message.