Celebrity Endorsements are the worst. Why do so many companies use them?

If you spent any amount of time in front of a screen today you probably saw a celebrity endorsement. Chances are you were also completely unaffected by it. While celebrity endorsements are an extremely popular advertising technique, they are also generally ineffective.

This should feel a little odd. Advertisements with celebrities are more likely to be shared or talked about so they’re likely to have increased viewership. So what do they do wrong? Well, actually, there are a few of things. Advertisements with celebrity endorsements do achieve higher engagement than comparable non-celeb ads, but they aren’t any better at helping consumers remember the brand. Basically, celebrities draw attention to the ads, but they typically draw attention away from the product and to themselves. This is known as the ‘video vampire’ effect.

Another psychological consideration is that consumers attribute the celebrity’s endorsement to the money paid to the celebrity instead of to underlying qualities in the product itself. One assumes this is particularly true when celebrities and brands occupy very different positions in our brains and that is often the case. Rob Lowe and DirecTv, Ozzy Osbourne and I Can’t Believe It’s Not Butter!, and Penelope Cruz and Nintendo are all good examples of this. Ozzy Osbourne in particular was probably chosen because the randomness would generate buzz, but since he doesn’t share many attributes with the brand that buzz probably didn’t convert into long-term product recall or improved sales.

The inherent “otherness” conveyed by celebrity is another drag on success. Social proof states that in uncertain situations people pay attention to others who are similar to themselves for tips on how to behave. That’s why brands often include actors in their commercials who are similar to their target market. When a celebrity and a brand (or the brand’s target audience) are mismatched it feeds the wrong information to potential consumers. Think about that in the context of Gameboy and Penelope Cruz. Gameboy’s target audience is presumably young boys. Ignoring her celebrity, the message Nintendo sends by using her is that the product is for middle-aged women. That, in itself, is mistaken, but if you add in the layer that she is a celebrity you have to take the mismatch one step further.

This isn’t to say that celebrity endorsements can’t be effective. In fact, that extra step can be a good thing for aspirational marketing. But, as in Nintendo’s case, brands often make a poor choice of endorser and the advertisements are more expensive. A more careful selection of celebrity (one who doesn’t endorse anything else, will stick around for a long time, and who closely matches the existing brand identity) will help these campaigns. Audi’s Grandpa Bode commercial is one I’ve always liked because it intimately relates the traits we identify him with to the product benefits conveyed by the advertisement.

Still, even with those changes, it might not be worth it for most brands. After all, as I’ve previously written, brand mascots (e.g. Verizon’s “can you hear me now” guy, the Geico gecko, or Flo from Progressive) play a similar role for businesses and they don’t present the same worries about potential legal trouble. It’s also worth noting that non-celebrity endorsements (AKA testimonials) are effective, and cheaper than celebrity endorsements.

In summary: Celebrity endorsements are dumb, but don’t expect them to go away any time soon.