The raw data on effective advertising

What works, what doesn't and why

by Paul Paquet

Marketing has compiled an array of research about the ads that don't work as opposed to those ads that do.

List plenty of benefits: Copernicus, a Westport, Connecticut company, found that 26 per cent of a test group wanted to buy a product described with one attribute or benefit. But 80 per cent would buy if eight benefits were listed.

Stress these four elements: Research Systems Corp. of Evansville, Indiana, found that four elements make an ad especially effective: 1) brand-differentiation, 2) new products or services, 3) convenience and 4) competitive comparison.

Don't annoy people: Figures from the US Advertising Research Foundation show that the more people like your ads, the more effective the ads are likely to be.

Find media that people like: According to a 1994 study by Simmons Market Research and Cahners Publishing, 52 per cent of readers are more likely to remember ads in publications that deeply interest them, while only 34 per cent remember ads in publications they read casually.

Run an ad 10 times: Conventional wisdom, dating from Herbert Krugman's 1972 article in the Journal of Advertising Research, suggested that you need to run an ad three times for it to have an effect. But data from Copernicus suggest that it takes 10 quick ad exposures to make an impact in a crowded media market.

Wait one month: Leslie Wood, of A-to-S Link, says that 20 per cent of consumers are more likely to buy within 24 hours of seeing an ad, whereas 11 per cent are more likely to buy within 28 days.

Combine ads and promotion: Copernicus also found that combining ads with promotion achieves 61-per-cent awareness. Relying mostly on promotion earns 25-per-cent awareness, while relying mainly on ads earns 42-per-cent awareness.

Cornerstone is a writing and editing firm that uses marketing and PR principles to create "words you can build your business on."

Call us from anywhere in the world for rush work.

Click here to exit .