Celebrity Endorsement Effectiveness

ConnectionsKnoll, J.; Matthes, J. · 2017Journal of the Academy of Marketing Science
Topicscelebrity endorsement·meta-analysis·match-up hypothesis·endorser credibility·advertising effectiveness·attitude toward object

Imagine you have a budget for a celebrity contract to launch a new skincare line. Before approving it, you weigh a famous model, an actor who fits the product, a quality seal and another brand's endorsement in the same category. Which endorsement should you fund?

Celebrity status alone has a near-zero average effect across consumer responses. Results vary with fit and endorsement format.

Data chart

Actors who fit the product and make no direct claim fare best

Predicted difference in how favorably people rate the endorsed product, celebrity ad versus the same product with no endorsement, in standard deviations (0 = no difference).

Zero means no differenceFits, no direct claim, female actor+.44Fits, no direct claim, male actor+.90Poor fit, no direct claim, femaleactor-.03 (n.s.)Poor fit, no direct claim, maleactor+.43Fits, direct claim, female actor-.11 (n.s.)Fits, direct claim, male actor+.35Poor fit, direct claim, female actor-.58Poor fit, direct claim, male actor-.12 (n.s.)
  • Fit and no direct claim work best for both sexes.
  • Female actors who fit and make no direct claim still score above no endorsement.

Predicted product attitudes are most favorable for actors who fit the product and make no direct claim, for both sexes.

Table 5, p. 66: predictions for an actor endorsing an unfamiliar product once, compared with no endorsement of the same product. These are model predictions, not observed averages. Standard deviations describe the gap relative to the spread in people's ratings. Exact values, female then male: fits, no direct claim, .44 and .90; poor fit, no direct claim, -.03 and .43; fits, direct claim, -.11 and .35; poor fit, direct claim, -.58 and -.12. For size, use the absolute gap: under .30 is small, .50 medium and above .67 large (p. 69).

Action guide

  1. Do not approve spending on celebrity status alone.Back your endorser choice with market research comparing the celebrity with quality seals, awards and same-category brand endorsements.
  2. Require your team to document celebrity fit.Shortlist actors ahead of models, musicians and television hosts, using a male actor as the benchmark in your market research.
  3. Favor appearances without a direct endorsement claim.Direct claims are linked to less favorable product attitudes, a medium to large gap.
  4. Prioritize celebrity support for new or unfamiliar products.Expect less lift in attitudes toward products consumers already know.

Evidence

  • Across all responses and comparison groups, the average celebrity advantage is .04 standard deviations, not reliably different from zero. Awareness, ad attitudes and behavioral intentions show no reliable average lift. The best predicted combination is a male actor who fits the product and makes no direct claim, while others can be strongly negative.
  • Compared with no endorsement, people rate products with a celebrity endorsement .24 standard deviations more favorably, a small difference. Unfamiliar products gain a little more than familiar ones.
  • Across studies, celebrity fit is linked to better product attitudes and behavioral intentions. On product attitudes, good fit beats poor fit by .47 standard deviations, after allowing for endorser sex and type, claim format, frequency, product familiarity and comparison group.
  • Across studies, appearing without a direct claim is linked to better product attitudes and behavioral intentions. A direct claim is linked to product attitudes .55 standard deviations lower, after allowing for fit and the same other factors.
  • Male endorsers outperform female endorsers on attitudes and behavioral intentions, a medium to large gap. Actors outperform models, musicians and television hosts on product attitudes, a small to medium gap. The gap for television hosts is reliable only before allowing for product familiarity.
  • Quality seals and awards are linked to better product attitudes than celebrity endorsements, and so, less certainly, are same-category brand endorsements. Against a quality seal or award, people rate the celebrity-endorsed product .44 standard deviations less favorably.

Key takeaway

Celebrity endorsement results are linked to fit and claim format, not celebrity status alone.

Limitations

The studies mostly cover U.S. print ads and immediate attitudes and intentions, not sales or observed behavior. Each factor's link comes from comparing studies, not from testing it within one experiment.

Source

Knoll, J., & Matthes, J. (2017). The effectiveness of celebrity endorsements: a meta-analysis. Journal of the Academy of Marketing Science. https://doi.org/10.1007/s11747-016-0503-8

https://doi.org/10.1007/s11747-016-0503-8