Caples on Testing

AN EXCERPT FROM: John Caples, 50 Things I’ve Learned in 50 Years

The key to success in advertising (maximum sales per dollar) lies in perpetual testing of all variables.

Over the years, many methods for testing copy have been devised. Opinion tests, readership tests, eye camera tests, pupilometer tests, recall tests, comprehension tests, c oupon tests, inquiry tests, attitude tests, etc. Most of these tests produced useful information.

Here is a simple test. When you write a piece of copy, put i t aside and read it over the next day. You will almost always be able to improve it.

Another simple method is to ask somebody to read your copy aloud. If he stumbles over a sentence, say to yourself: “That’s not his fault. It’s my fault. I must make the sentence better.”

If you want to get an associate’s opinion of an ad you wrote, don’t show him just one ad. Chances are he will try to please you b y saying: “It’s good.” That gets you nowhere. Show him two ads and say, “Which is better?”

Testing ads by asking people for their opinion is helpful. However, i t can be misleading. Many will not vote for all-type ads. Most believe that an ad is not good unless it has a picture. This is not so. Some of the best-pulling mail order ads have had no pictures.

In an opinion test, people hesitate to reveal their selfish motives. For example, in an opinion test of life insurance ads, an ad with the headline “What would become of your wife if something happened to you?” outpulled an ad with the headline “To men who want to quit work some day.” When these ads were subjected to a mail order sales test, the results were reversed.

Do not discard opinion testing because it is sometimes inaccurate. Opinion testing has one big advantage over mail order tests. You can ask the respondents why they voted for a certain ad. You can find out if the copy is understood or misunderstood. You cannot do these things in a mail order test.

The best tests, if properly handled, are sales tests. Mail order advertisers have an advantage in this respect. Every mail order ad is a sales test. In mail order, you can test copy, media, position in media, and season – all by sales results. Hence, mail order advertisers know a great deal about the realities of advertising. Much of this knowledge is applicable in those forms of advertising which cannot be accurately tested.

The most accurate test is a mail order split-run test where two ads – Ad A and Ad B – are tested under identical conditions. Many publications offer split-run copy testing. They do this by splitting the press run. Ad A runs in half the circulation.Ad B runs in the other half of the circulation – same issue, same page, same position on the page. If a news dealer has 100 copies of the publication, 50 copies will contain Ad A and 50 copies will contain Ad B. Thus each ad has an equal chance to get results.

Testing copy is fun, exciting, rewarding. I recall working on ads for a finance company that offered small loans. Several of us wrote ads and we tested them in newspapers by counting phone calls from prospects. For example, one ad would say “Telephone this number and ask for Miss Smith.” Another ad would say “Ask for Miss Miller,” and so on. Thus we could tell exactly how much business each ad brought in. Then each copywriter would bet a dollar that his ad would win. Testing copy became a game we all enjoyed. It was as thrilling as betting on a horse race. We learned a lot. And the client benefited.

P.S. If you’d like to learn even more about how to write copy that sells, I recommend you join the private Copywriting Code membership site. It’s packed with valuable copywriting lessons, and more are being added all the time. Click here to check it out.

One Comments

  1. Ryan: “Here is a simple test. When you write a piece of copy, put i t aside and read it over the next day. You will almost always be able to improve it.”

    Yes! Yes! Yes! Cannot recommend this more strongly myself. Almost without exception, I will always find something to tweak that should have jumped out during the first read through right after writing it, but didn’t.