07 Mar 2019 The engine that finds pieces of evidence for your marketing claims

Let me start with a question: what do these three sentences have in common?

  • “Lactobacillus paracasei supports a healthy intestinal flora.”
  • “Vitamine B3 activates the scalp microcirculation.”
  • “Sugar-free chewing gum is beneficial to dental health.”

I bet that if you read the first one on a yoghurt cup, you will be more inclined to buy it. The second one will probably make you go for this bottle of shampoo over another one that does nothing to your scalp (except clean it). Maybe the last one will raise your eyebrows and get you thinking, “Really?”

Now to the answer: these three sentences are all marketing claims that have been rejected by the European Commission (EC). The EC has ruled that these claims should not be used to promote a product because there’s no way to prove they are true. A rare occurrence? Not in the least: the EU register of nutrition and health claims contains a total of 2,337 entries, only 261 of which are authorized (as of January 2019). Almost 90% of the registered claims have been rejected.

From the consumer’s point of view, this is good. Thanks to the vigilance of the EU, you can be sure that what’s written on the packaging is really inside the product. In the US, consumers are protected from misleading marketing claims by the Federal Trade Commission (FTC). The FTC makes sure that if your breakfast cereal is labeled “all natural”, it really is.

From a company’s standpoint, however, this mountain of rejected marketing claims means a lot of money.

In the US, litigation settlements with the FTC can reach millions of dollars (in the case of the not-so-natural “all natural” breakfast cereals, one company agreed to a $5 million settlement just in California). For consumer goods companies, it has become vital to check the veracity of marketing claims BEFORE a competitor or a consumer rights protection association files a complaint.

How does this actually work? I imagine a modern conference room, large bay windows, a view over a sea of skyscrapers. On one side of the room, the creative team with their funny T-shirts and fancy sneakers; on the other side, the legal team, stiff as brooms in their grey suits. If you bend an ear, you might hear music by Ennio Morricone. The duel is about to begin. The marketers want people to buy their products; they must convince them that their product is better, healthier, more effective. To do so, they come up with the sexiest tag lines ever: an energy drink that “gives you wings” or walking shoes that “increase calorie burn”. But the lawyers on the other side of the room don’t care about creativity. Millions in litigation costs are at stake. They need to make sure that the creative folks are not stretching the truth too far.

Who will win the battle? The team with the better argument, of course—the team that can put on the table a clinical study or research report that confirms or denies the marketing claims.

But finding a piece of evidence requires sifting through thousands of scientific publications, clinical studies and consumer tests.  A long, tedious and rather unrewarding task, as you can imagine. A task that one would gladly delegate to a robot. A task that Cortical.io Semantic Search engine does without batting an eyelid (if it had one).

There are a couple of challenges in automating the marketing claim verification process though.

For example, the scientific vocabulary used in research studies strongly differs from the marketing jargon used to describe a product to consumers. The scientific information relevant to a product is also not often formulated in words, but in recipe formulations. Common search engines fail to overcome these semantic hurdles.

Recently, Cortical.io Semantic Search engine has been successfully trained to enable search using queries formulated as marketing claims to identify related scientific information. Users have the possibility to filter search results based on common formula names, formula codes or a combination of formula ingredients. The engine was trained very quickly based on annotations by the firm’s subject-matter experts on only a handful of sample clinical reports. Cortical.io Semantic Search engine understands the meaning of marketing claims and matches those claims with research studies even if they do not use the same vocabulary. As a result, new sources of evidence within scientific literature could be discovered.

With Cortical.io technology, that company is on the verge of automating its marketing claim verification process. The engine also enables the reallocation of expensive subject-matter experts’ time to more demanding tasks—a win-win situation for both the employee who can spend time on more rewarding tasks and the employer who can make better use of resources and improve motivation and efficiency.

With intelligent machines like Cortical.io Semantic Search engine, complying with regulatory requirements does not have to be a hassle anymore–whether it is fun is another question!

Author: Marie-Pierre Garnier,
VP Marketing & Communications
back to opinions overview