Jan Hofmeyr outlines the steps we need to take to make research more predictive.
Tracking surveys are in trouble, but their plummeting response rates, faulty answers and flawed analysis can easily be solved by the disciplined application of three core principles.
Imagine two people, a man and a woman. On the 37th question in a tracker survey, they are asked how likely they are to take a holiday in Thailand next year. The man answers “very”. The woman answers “not very”. We conclude that Thailand is likely to have one visitor; and it’ll be a man. We advise our client to prepare for a man; and send the man tourist information.