Thursday, July 13, 2017

Making the Most of Online Research

Online research gets a bad rap.  Some of the arguments I have heard against it:  "The sample size is too small."  "It's unreliable."  "Only weirdos take online music surveys."

On the flip side, it's cost effective.  It has a crazy fast turn around, and for many stations, it's the only research tool they have.

So, how can you make the most out of your online research?  Here are some strategies I have used which have resulted in, what I believe, is a pretty reliable tool to help make my station is as effective as possible.

1)  Split your database:  When people sign up for your online research, they answer many questions about their age, where they live and sometimes even their interests.  Nose around in all that data and figure out who you trust the most.  Put them in one database.  Everyone else goes in the other.  I run two surveys simultaneously.  Every single time.  One survey group is my target.  The other is everyone else.  I look at data from both groups, but weigh the opinion of the target group more.

2)  Know who is in  your database:  Theoretically, the people who are signed up to take your music surveys are passionate about your station.  Hopefully they come to your events.  Meet them and get to know their names.  If that doesn't happen naturally, plan an event for your database.  Some sort of "Thank You" gathering with some cost-effective food and goodie bags.  This is where you might actually see some weirdos.  Make sure you have them in the right database.  (See above.)

3)  Massage your database:  I fiddle in my database every couple months.  As new people join, I move some to my target as needed.  As people age, I move them out of the target database.  Recently, I heard some programmers talking about testing W25-49 instead of W25-54.  The change in my results was significant.  The result is a fresher, yet still hit-based, sound. Give it a try.  If that doesn't work for you, move the 49-54 back in next time.

4)  Reward your database:  Start every message with "Thank You."  End every message with "We appreciate you!"  Let your database know they are important.  Then, give them stuff.  Random stuff is great.  Exclusive stuff is better.  Get them concert ticket pre-sale codes.  And on your next email, tell everyone who won the last time.  The goal here is not to bribe people.  The goal is to create buzz.  It gives your air personalities a reason to talk about your survey on the air, and reminds people to take it.  A giveaway can be especially helpful during those times of year when response slows down.  

5)  Remember you database is a tool:  With any home improvement job, you depend on a variety of tools to get the job done.  Programming a radio station is like that.  Online research, combined with all the other information we have at our fingertips, gives us a clear view of trends that are emerging with our audiences.  I don't live and die by my online research, but when it confirms what I already believe, I make a change and don't look back.

6)  Throw out bad surveys:  Once in awhile, you get back a wacky survey.  It happens.  By testing a variety of songs, you can easily tell if your data is good.  If proven hits suddenly take a turn and new music is testing like a rockstar, that would be weird.  In the four years I have been using online testing, I have tossed a few.  Trust your gut.  If the data doesn't quite look right, don't use it. If you have a few bad surveys in a row, massage your database.  (See above.)

7)  Keep adding new members:  People will eventually stop taking your surveys.  They get bored.  Their email address changes.  They get offended with your station.  Keep putting new people into the pipe.  So how many people do you need to have to make your data reliable?  As many as possible.  Once I asked a trusted consultant how many people had to show up at an auditorium test to have it count.  His answer was 35, if they are the right people.  This is where running two databases is helpful.  (See above.)  You will get more reliable results from a small group of the right people than a large group of the wrong people.

How do you get the most out of your online testing?  I'd love to learn ways to improve what I do.  Hopefully I've given you some food for thought.  As always, I appreciate your feedback.

Terese Main is Assistant Program Director/Music Director/Morning Co-host at Family Life, a network of nearly 70 Christian radio stations across New York and Pennsylvania (  She is also a freelance on-air personality in Kansas, Nevada, Arizona and Texas and also provides imaging voicework for stations in Florida, California, Minnesota and Wisconsin (