The C.A.B. system really did the same damn thing as Hooper except they called after the program ended! There were some shortcomings from the start at the Cooperative Analysis of Broadcasting Company. But advertising research was in it's infancy then. It was only in 1908 that Walter Dill Scott, a Northwestern professor published "the Psychology of Advertising." It's here.
The result was that confidence in radio ratings was iffy and listenership was a bit of a mystery and more importantly none of the polling methods were proven. The audience measurement process has varied greatly since it's inception, and although It's roots lie in that first book by Dill-Scott, personally I blame Archibald Maddock Crossley.
In 1918 Crossley was commissioned to form a surveying team for the city of Philadelphia. Early goals were to determine the actual reach of stations. QSL cards were useful, but not broadly applicable. mostly they wanted to know if local affiliates were running their network ads. (Note that's a concern that continues to this very day.) In 1926 he founded the New York market research firm Crossley, Inc and by 1929 Crossley was doing phone surveys. He produced the first radio ratings ever. It was a big enough deal that he wrote a book on the concept called "Watch your Selling Dollar!"
Two big advertisers at the Association of National advertisers wanted to get more empirical. The big push came from Kodak and Davis baking powder. In 1930 the Association of National Advertisers (ANA) joined forces with Crossley to get the job done. They met on February 7th at the Yale Club in New York, he agreed to undertake the task with their endorsement. The Cooperative Analysis of Broadcasting Company was founded. More here.
In the first year 49 advertisers subscribed. The wave had begun. Crossley used random numbers from telephone books calling people in about thirty. Then his interviewer would ask about the radio programs they had listened to the day before. This method became known as the recall method. This went on for 16 years. They had a monopoly until 1934 when Hooper was founded, after that it was a slow slide downhill. the first Hooper reports differed from the existing Crosley reports. In the face on continual criticism over methodology, he closed up shop and sold to Hooper.
He continued to do political polls until 1952, retiring in 1962. In 1970 he won an AAPOR award. He died in 1985. Time magazine wrote an epitaph. His grandson Josheph Crossley II founded the Crossley machine Company.