Testing is essential
if you want to remain profitable

---

If you don’t invest enough time in testing you will definitely make some expensive mistakes. Even if you have to rent a mailing list of 5,000 names, you don’t have to send to all of them at once. Send to just 200. If it doesn’t work, send a different letter and offer to the next 200. Keep going until you hit something that works. Then you can ramp up and mail the rest of the list.

A time-honoured method is to increase 5-fold each time you test. In other words, if 200 gives a good response, try 1,000. If that works, then try 5,000. If that works then gor for the 25,000. This depends on the size of your target market, of course. Whatever you do, don’t just go from 200 to 25,000. Two hundred is not a big enough sample to give confidence that you’ll get the same percentage response for higher numbers.

Take the same approach with your advertising. Place small ads in a publication before taking a bigger ad. Measure the success of a large ad by mail or leaflet before committing to running it in a magazine or newspaper.

A Low-cost Action Plan for Testing

1. Look at any mailshot you’re planning. What’s the minimum number you could send to get a meaningful response? Make sure you don’t send any more than that. If it’s successful, multiply by 5 the next time, and repeat until you’ve hit all your target audience.

2. Try sending different letters to different samples to gauge the difference in response. Use the best letter when you ramp up to the next level.

3. Make sure you pre-test your ads. Place small ads to make sure the readership is right for you. Test your larger ads by using them as mailers or leaflets - it’s much cheaper that way. Again you can try different versions of the ad on different sample groups, then use the best one when you run it for real.